Regulation & The Evolution of The Global Telco Industries

Regulation and the Evolution of the Global Telecommunications Industry

To my mother and the memory of my father who provided encouragement

and guidance in ways they never imagined. Also to my colleague and best

friend, Gary Lumsdem, who started his successful new business by taking

advantage of lucrative opportunities created by the new regulations of

telecommunications industry.

Anastassios Gentzoglanis

To my good colleagues at the Center for Communication, Media and

Information Technologies (CMI) at Aalborg University in Copenhagen

with whom I have cooperated for many years.

Anders Henten

Regulation and the Evolution of the Global Telecommunications Industry

Edited by

Anastassios Gentzoglanis

University of Sherbrooke, Canada

and

Anders Henten

Aalborg University, Denmark

Edward ElgarCheltenham, UK • Northampton, MA, USA

© Anastassios Gentzoglanis and Anders Henten 2010

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher.

Published byEdward Elgar Publishing LimitedThe Lypiatts15 Lansdown RoadCheltenhamGlos GL50 2JAUK

Edward Elgar Publishing, Inc.William Pratt House9 Dewey CourtNorthamptonMassachusetts 01060USA

A catalogue record for this bookis available from the British Library

Library of Congress Control Number: 2009940969

ISBN 978 1 84844 588 8

Printed and bound by MPG Books Group, UK

02

v

Contents

List of contributors vii

Foreword by Stephen Schmidt xvi

Preface by Erik Bohlin xvii

Evolving technologies, competition and the new role of regulation:

introduction and synopsis of the book 1

Anastassios Gentzoglanis and Anders Henten

PART I REGULATION VERSUS INVESTMENT:

THE BALANCE BETWEEN STATIC AND

DYNAMIC EFFICIENCIES AND THE MAIN

ISSUES OF REGULATORY POLICY

1 Investment in broadband technologies and the role of

regulation 21

Anastassios Gentzoglanis and Elias Aravantinos

2 Intermodal telecommunications competition: implications

for regulation of wholesale services 43

William E. Taylor

3 Access regulation versus infrastructure investment:

important lessons from Australia 63

Martyn Taylor

4 Behavioral economics and telecommunications policy 83

Patrick Xavier and Dimitri Ypsilanti

PART II TECHNOLOGY CONVERGENCE AND

THE FUTURE ROLE OF COMPETITION AND

REGULATION

5 The measure and regulation of competition in

telecommunications markets 109

Marcel Boyer

6 Preventing harm in telecommunications regulation: a new matrix

of principles and rules within the ex ante versus ex post debate 128

Kenneth Jull and Stephen Schmidt

vi Regulation and the evolution of the global telecoms industry

7 The transformation of telecoms industry structure:

an event study 149

Olaf Rieck

PART III ACCESS REGULATION AND PERFORMANCE

IN THE DEPLOYMENT OF NGNS:

INTERNATIONAL EXPERIENCES

8 From the pursuit of effi ciency to the pursuit of competition in

New Zealand’s evolving telecommunications market 167

Bronwyn Howell

9 International regulatory comparisons: the evolution of

IP- based fi ber 188

Scott Marcus and Dieter Elixmann

PART IV STRUCTURAL SEPARATION AND

REGULATION OF THE

TELECOMMUNICATIONS INDUSTRY

10 Diff usion of broadband Internet and structural separation 211

Arata Kamino and Hidenori Fuke

11 Implementing functional separation in fi xed

telecommunications markets: the UK experience 233

Peter Curwen and Jason Whalley

12 Effi ciency and sustainability of network neutrality proposals 253

Toshiya Jitsuzumi

PART V INTERDEPENDENT INNOVATIONS AND

REGULATORY POLICIES: MOBILE NETWORK

DEPLOYMENT AND MOBILE INTERNET

DEVELOPMENTS

13 Interdependent innovation in telecommunications: risk,

standardization and regulation 275

Bruno Basalisco, Andy Reid and Paul Richards

14 Next generation mobile networks deployment and regulation

in the European Union 301

Claudio Feijóo, Sergio Ramos and José- Luis Gómez- Barroso

15 Mobile Internet developments in Europe, East Asia and the US 317

Morten Falch, Anders Henten and Karsten Vandrup

Index 341

vii

Contributors

Elias Aravantinos has been Project Manager at the Columbia Institute

for Tele- Information on the Ultrabroadband project since 2004. He is

also the Managing Director of ExelixisNet which specializes in high-

technology strategies and new media products. His clients are carri-

ers, governments, academic institutes and media companies in North

America and Europe. From 2006 till 2008, Mr Aravantinos worked for

Queens College and the Metropolitan College of New York, teaching

several business and technology courses at graduate and undergradu-

ate level. He is completing his PhD thesis in Technology Management

at Stevens Institute of Technology, USA with a specialization in

Telecommunications. His research interests are new technology projec-

tions, strategies and business development, with a focus on 4G strategies.

He is a frequent speaker at industry and academic events, with numerous

presentations and publications. Mr Aravantinos holds a BSc degree in

Electrical Engineering and Computer Technology from the University of

Patras, Greece and an MBA in Information Technology from Oklahoma

City University, USA.

Bruno Basalisco is a PhD candidate in Economics at Imperial College

Business School, London. He conducts research in the fi eld of industrial

economics, focusing on innovation, regulation, platforms and business

models in the information and communications and network industries in

general. His work has been presented at several international conferences

and workshops in the fi eld of industrial organization, technology strategy

and telecommunications policy, and he has served as referee for leading

information and communication technology (ICT) economics and strat-

egy journals. Bruno holds an MSc in Science and Technology Policy from

SPRU, University of Sussex and an MA in Economics from Sant’Anna

School of Advanced Studies and University of Pisa. He has also briefl y

worked for NERA Economic Consulting and for Ofcom.

Erik Bohlin is currently Head and Professor in Technology Assessment

in the Division of Technology and Society, Department of Technology

Management and Economics at Chalmers University of Technology. He

has published in a number of areas relating to the information society

– policy, strategy, and management. He is Chair of the International

viii Regulation and the evolution of the global telecoms industry

Telecommunications Society and Chief Editor of Telecommunications

Policy. He served as Special Advisor to IT and Media Commissioner

Viviane Reding of the European Commission in 2008–2009. He obtained

his graduate degree in Business Administration and Economics at

the Stockholm School of Economics (1987) and his PhD at Chalmers

University of Technology (1995).

Marcel Boyer holds a PhD in Economics from Carnegie- Mellon University.

He has taught economics at York University (1971–73), UQÀM (1973–74)

and at the University of Montreal (1974–2008). He held the Bell Canada

Chair in Industrial Economics in the Department of Economics at the

University of Montreal (2003–08) and the Jarislowsky–SSHRC–NSERC

(Social Sciences and Humanities Research Council; Natural Sciences

and Engineering Research Council of Canada) Chair in Technology and

International Competition at l’École Polytechnique de Montréal (1993–

2000). He is presently Emeritus Professor of Economics at the Université de

Montréal, Fellow of the C.D. Howe Institute, Centre interuniversitaire de

recherche en analyse des organisations (CIRANO) and Centre interuniversi-

taire de recherche en économie quantitative (CIREQ), Academic Affi liate of

The Analysis Group, Senior Economist at the Montreal Economic Institute,

member of the board of the Agency for Public–Private Partnerships of

Québec, member of the Industry Canada Advisory Committee on Business

Strategies and innovation, member of the Governance Committee of the

Sustainable Development and Socially Responsible Investment Chairs

(École polytechnique de Paris and Université de Toulouse). Marcel Boyer

has received numerous prizes for excellence in research. He is author or

co- author of over 250 scientifi c articles and papers and public and private

reports. Marcel Boyer has acted as an expert economist on behalf of several

national and international corporations and government organizations, and

has testifi ed as an expert witness before various organizations and tribunals.

Peter Curwen is Visiting Professor of Telecommunications Strategy at the

Strathclyde Business School, Glasgow and also researches and publishes

on a private basis. He was previously Professor of Business Economics at

Sheffi eld Hallam University. His primary research interests concerns the

manner in which a rapidly changing environment aff ects the structure of

the mobile telecommunications industry and its strategic implications for

major companies in that sector. He has published three books on telecom-

munications, including Telecommunications Strategy: Cases, Theory and

Applications (Routledge, 2004) with Jason Whalley.

Dieter Elixmann is head of the Corporate Strategies research group

at Wissenschaftliches Institut fur Kommunikationsdienste (WIK), Bad

Contributors ix

Honnef, Germany. He has been with WIK since 1985. He is also head

of the Telecommunications Markets research group. His recent research

focuses on econometric estimation of production structures of Deutsche

Telekom, demand analysis in the telco sector, internationalization and

globalization in the telco sector, and multimedia business strategies.

Morten Falch is Associate Professor at the Center for Communication,

Media and Information Technologies (CMI) located at Copenhagen

Institute of Technology, Aalborg University, Denmark. He holds a

Bachelors degree in Mathematics, a Masters degree in Economics and

a PhD. Since 1988 his research has been specialized in socio- economic

issues related to information and communication technologies. This

includes economic analysis of applications and telecommunication net-

works and services (for example cost analysis of telecom networks),

e- government, regulation of the telecom sector (in particular regulation

of interconnection), information and communication technology (ICT)

and industry policy, the role of competition in innovation of new serv-

ices and spectrum management. He has participated in many European

Union (EU) funded research projects in the telematics area. He has also

conducted a large number of consultancies for national and international

organizations such as the International Telecommunication Union (ITU),

the United Nations Conference Trade and Development (UNCTAD), the

World Bank and the national telecom agencies in Denmark, Norway and

Sweden.

Claudio Feijóo is Professor at Universidad Politécnica de Madrid, Spain

where he researches on the future socio- economic impact of emerging

information society technologies. He recently returned from a highly

rewarding two years as a visiting researcher at the Institute for Prospective

Technological Studies of the Joint Research Centre of the European

Commission. Since 1993 he has been involved in telecommunications and

information society development from the academic, the practical imple-

mentation and the public administration perspectives.

Hidenori Fuke has been Professor of Info- Communications Industry at the

Faculty of Global Media Studies, Komazawa University in Tokyo, Japan

since 2007. He has done extensive research on the info- communications

industry and regulatory policy in the industry as a member of the board

of directors at InfoCom Research and Professor at Kansai University in

Osaka. He received his BA in Economics from Tokyo University, and M

Litt. from Glasgow University. He also received a PhD in International

Public Policy from Osaka University. His major areas of specialization

include the economic analysis of the telecommunications industry and

x Regulation and the evolution of the global telecoms industry

regulatory policy. He has written many papers and books in this area in

both Japanese and English.

Anastassios Gentzoglanis holds a PhD degree from McGill University,

Canada and is a full Professor of Economics and Finance at the University

of Sherbrooke. He teaches and conducts research in the area of regulatory

economics. He is frequently invited to teach and work on research projects

related to the economics of regulation and new technologies in Europe,

Asia, Central and Latin America and Africa. He has also received many

funded research grants from both domestic and international organiza-

tions, published in a number of high- quality scholarly journals and partici-

pated in many international conferences. In June 2008, Dr Gentzoglanis

organized and chaired the 17th Biennial conference of the International

Telecommunications Society (ITS) in Montreal. He is also director of the

Centre for the Study of Regulatory Economics and Finance (CEREF) and

he organizes courses for executives in the area of regulation of public utili-

ties for the Francophonie.

Toshiya Jitsuzumi is a Professor of Industrial Policy at the Faculty of

Economics, Kyushu University, Japan. He has earned an LLB from the

University of Tokyo, an MBA from New York University, and a DSc

from Waseda University, Japan. Prior to starting his academic career

in 2000, Professor Jitsuzumi had served for 15 years at the Ministry of

Post and Telecommunications (now Ministry of Internal Aff airs and

Communication), Japan. The research topics of his interest include public

economics and communication and Internet economics. During 2007–08, he

was a visiting scholar at the Columbia Institute for Tele- Information (CITI),

Columbia University, as an Abe Fellow sponsored by the Social Science

Research Council/Japan Foundation Center for Global Partnership. His

research work has appeared in various journals, including Telecommunications

Policy, Foresight and Socio- Economic Planning Sciences. He is a member of

several academic societies, including the International Telecommunications

Society and the Regional Science Association International.

José- Luis Gómez- Barroso is Associate Professor at the Universidad

Nacional de Educación a Distancia (UNED), Spain. He holds a PhD in

Economics, as well as a Masters in Telecommunication Engineering and

a Masters in Law. His teaching and research interests lie in the evolution

of the information society and electronic communications under the triple

perspective of technology, economy and regulation, Dr Gómez- Barroso

has contributed signifi cantly in the fi eld of telecommunications by author-

ing many papers in academic journals and international conferences on

the subject.

Contributors xi

Anders Henten is Associate Professor at the Center for Communication,

Media and Information technologies (CMI) at Aalborg University,

Denmark. He is a graduate in Communications and International

Development Studies from Roskilde University in Denmark and holds

a PhD from the Technical University of Denmark. His main areas of

research are service innovation and internationalization, regulation of

communications, standardization, and socio- economic implications of

information and communication technologies, including e- business and

business modeling. Anders Henten has worked professionally in the areas

of communications economy and policy for more than 20 years. He has

participated in numerous research projects fi nanced by the European

Community, the Nordic Council of Ministers, Danish research councils

and ministries, and in consultancies, fi nanced by the World Bank, the

United Nations Conference on Trade and Development (UNCTAD), the

International Telecommunication Union (ITU), Danish ministries, and

so on. He has published nationally and internationally – more than 200

academic publications in international journals, books, anthologies, con-

ference proceedings, and so on.

Bronwyn Howell is General Manager of the New Zealand Institute for the

Study of Competition and Regulation, and a faculty member of Victoria

Management School, Victoria University of Wellington, New Zealand.

She teaches and researches in the areas of institutional economics, the

information economy and regulation, with specifi c interests in the devel-

opment, implementation and performance of institutions and policies

governing the operation of information technology and telecommunica-

tions markets.

Kenneth Jull practices at Baker & McKenzie LLP, Toronto offi ce, in the

area of risk management strategies to promote regulatory and corpo-

rate compliance. Mr Jull is the co- author with Justice Todd Archibald

and Professor Kent Roach of Regulatory and Corporate Liability: From

Due Diligence to Risk Management. He is an Adjunct Professor at the

University of Toronto, Faculty of Law where he teaches ‘Financial

Crimes’. Mr Jull has an appointment to the Faculty of Graduate Studies at

Osgoode Hall Law School, where he is the Director of the part- time LLM

specializing in civil litigation and dispute resolution.

Arata Kamino has been an Executive Researcher of the global information

and communication technology (ICT) industry study group at InfoCom

Research, Inc. in Tokyo, Japan since 1995. He has been responsible for

comprehensive analysis of ICT industry using method of comparative

study between major overseas countries and Japan. His main interests

xii Regulation and the evolution of the global telecoms industry

cover policies, regulations, market structure and corporate strategies in

the fi eld of domestic and global ICT. He received his BA in Economics

from Tokyo University. He has written many papers for books, academic

journals and magazines.

Scott Marcus is a Senior Consultant for WIK- Consult GmbH, Germany,

a research institute in economics and regulatory policy for network indus-

tries. Previously, he served as Senior Advisor for Internet Technology for

the US Federal Communication Commission (FCC). Prior to that, he

was the Chief Technology Offi cer (CTO) of Genuity, Inc. In 2004, Scott

was attached to the European Commission (DG INFSO – Directorate

General for information Society and Media) as a Transatlantic Fellow

of the German Marshall Fund of the United States. Scott is also a newly

appointed Fellow of GLOCOM (the Center for Global Communications,

a research institute of the International University of Japan), and a

Visiting Fellow of the University of Southern California’s Center for

Communication Law and Policy. He is co- editor for public policy and

regulation for IEEE Communications Magazine. He served on the board

of the American Registry of Internet Numbers (ARIN) from 2000 to 2002,

and on the Meetings and Conference Board of the Institute of Electrical

and Electronics Engineers (IEEE) Communications Society from 2001

to 2005, and as Chair of the IEEE Committee on Network Operations

and Management (CNOM). He is a Senior Member of the IEEE. He is

the author of numerous papers and of a book on data network design:

Designing Wide Area Networks and Internetworks: A Practical Guide

(Addison Wesley, 1999).

Sergio Ramos gained his PhD degree in Telecommunications Engineering

from the Universidad Politécnica de Madrid, and he holds an MBA

from the Stockholm School of Economics. He worked for the European

Commission as Resident Twinning Adviser of a European Commission

(EC) Twinning Project for the Public Utilities Commission of Latvia,

to design and monitor the transposition process of the European Union

(EU) framework into Latvian legislation. He is currently Deputy Director

at the Spanish Association of Telecommunication Network Operators

(REDTEL).

Andy Reid has the role of Chief Network Services Strategist at British

Telecom (BT). Prior to this position he has worked in technical as

well as product marketing roles in BT. His professional interests and

expertise cover telecommunications technologies, network architecture,

service pricing, and links to regulatory and competition law microeco-

nomic analysis. He made major contributions to the development of

Contributors xiii

the telecommunications transmission technology including SONET/SDH

(synchronous optical networking/synchronous digital hierarchy) and more

recently on Ethernet standards, and has co- authored two major textbooks

in this area as well as numerous papers for professional standards bodies

and technical forums.

Paul Richards is a Regulatory Economist and has worked for British

Telecom (BT) since 1997. Prior to this position he worked in a variety

of private and public sector roles in the areas of economic modeling and

regulatory analysis.

Olaf Rieck is an Assistant Professor at Nanyang Business School,

Singapore. His research focuses on telecommunications economics and

policy, as well as on corporate strategy for telecommunications service

providers. Dr Rieck studied Industrial Engineering and Management

(Universität Karlsruhe/Germany, University of Auckland/New Zealand)

and Economics (University of British Columbia/Canada) and holds a PhD

in Business Administration (University of British Columbia). Previous

to his academic career he worked for Mercedes- Benz AG on strategic

information technology (IT) projects in the areas of telecommunications

and business process re- engineering. Dr Rieck teaches an introduction to

IT, telecommunications industry management, and mobile commerce to

undergraduate students and MBAs. He has also held executive teaching

assignments with Korea Telecom and LG.

Stephen Schmidt is Chief Regulatory Legal Counsel at TELUS

Communications Company. His responsibilities include regulatory,

legal and policy submissions to the Canadian Radio- television and

Telecommunications Commission (CRTC), the Federal Court of Appeal,

the Supreme Court of Canada, the Competition Bureau and other govern-

ment bodies. Prior to joining TELUS, he worked at AT&T Canada and

ACC TelEnterprises, with a focus on regulatory and legal matters respec-

tively. Mr Schmidt holds a BA (with Distinction) from the University of

Toronto and an LLB from the University of Manitoba. He is on the board

of directors of the International Telecommunications Society and lives in

Ottawa, Canada.

Martyn Taylor is a Partner in the competition and market regulation team

of Gilbert & Tobin, based in Sydney. He specializes in telecommunica-

tions law, antitrust law and international economic law. Martyn has been

advising on telecoms law and regulation since 1995, principally within the

Asia- Pacifi c region. His clients have included some of the world’s largest

technology and telecoms multinationals. Martyn has multiple postgradu-

ate qualifi cations in law, corporate fi nance and economics, including a

xiv Regulation and the evolution of the global telecoms industry

PhD. He has published extensively, including a book titled International

Competition Law (Cambridge University Press, 2006) and as a contribu-

tor to Merger Control Worldwide by Dabbah and Lasok (eds), Cambridge

University Press, 2008. He is currently the technology editor for the

Australian Business Law Review.

William E. Taylor is a Senior Vice- President at NERA Economic

Consulting, where he heads the Communications Practice and the Boston

offi ce. He received a BA in Economics from Harvard, an MA in Statistics

and a PhD in Economics from Berkeley. He taught economics, sta-

tistics, and econometrics at Cornell, Louvain and MIT and published

research in economics, econometrics and telecommunications policy at

Bell Laboratories and Bellcore. At NERA, he has worked and pub-

lished extensively in telecommunications economics on such issues as

access charges, costs, regulatory reform, productivity, competition policy,

mergers, vertical integration, interconnection pricing and antitrust issues.

He has testifi ed on telecommunications economics before numerous

domestic and foreign regulatory authorities, courts and legislative bodies.

Karsten Vandrup, Director at Litepoint Europe, is currently responsible

for Litepoint’s sites in Israel and Denmark. Prior to this position, Vandrup

was Vice- President of Engineering at Litepoint’s Headquarters in Silicon

Valley, California. Before joining Litepoint, Vandrup held several posi-

tions at Nokia including Senior Research Manager in the Nokia System

Research, Manager of Strategic Planning in the Nokia Research and

Technology Access (Espoo, Finland), Global R&D Cooperation Manager

in the Nokia Research and Education Policy Department at the Nokia

Head Offi ce (Copenhagen). As a part of the latter, Vandrup was a member

of the Career- Space Consortium. He started his career in Nokia as DSP

Design Engineer, later Engineering Manager, and holds an akademiin-

geniør degree in Telecommunication and Electronics from the Technical

University of Denmark, supplemented with studies at INSEAD, the

University of California, Los Angeles (UCLA) and the Swedish School of

Economics and Business Administration in Helsinki.

Jason Whalley is Reader in the Department of Management Science,

University of Strathclyde, Glasgow. His research interests include the

internationalization of mobile operators, as well as the development of

telecommunications policy in mountainous developing countries. In addi-

tion, his research also examines the diff usion of broadband and the regula-

tory measures taken to encourage its uptake.

Patrick Xavier is Director of Info- Comm Strategies and Adjunct Professor

of Communications Economics, Centre for Communications Economics

Contributors xv

and Electronic Markets, Curtin University of Technology, Australia. He

has served as consultant to a wide range of national and international

agencies, including the Organisation for Economic Co- operation and

Development (OECD), World Bank, Asia- Pacifi c Economic Cooperation

(APEC) and International Telecommunication of Union (ITU).

Dimitri Ypsilanti is head of the Information, Communications and

Consumer Policy Division in the Directorate of Science, Technology and

Industry at the OECD. He has specialized in telecommunication policy

and regulation.

xvi

Foreword

TELUS is very proud to have been the corporate host, along with the

Université de Sherbrooke as the academic host, of the 17th Biennial

Conference of the International Telecommunications Society (ITS): The

Changing Structure of the Telecommunications Industry and the New

Role of Regulation, which took place in June 2008, in Montreal, Canada.

TELUS invited the ITS conference to Canada because TELUS believes in

fostering high- quality analysis of important public policy issues aff ecting

telecommunications both in Canada and around the world.

More than 300 delegates from more than 40 countries came together for

four days in Montreal to share new ideas and approaches for dealing with

pressing issues in telecommunications, including technological change,

new sources of competition, new theoretical developments, and the ways

in which government policy can best respond to these issues.

This book fl ows directly from the ITS conference and eff ectively cap-

tures the character of the conference in microcosm. The 15 chapters

selected by Drs Gentzoglannis and Henten refl ect the rich, international

character and professional diversity of ITS and its members. The present

volume contains contributions from the academic, private sector and

government communities, providing a crucial window into key contem-

porary debates in telecommunications. I am confi dent that this volume

will inform and shape future debate, policy- making, business planning,

research and strategic thinking.

Stephen Schmidt

Chief Regulatory Legal Counsel

TELUS Communications

Ottawa, Canada

xvii

Preface

The International Telecommunications Society (ITS) convened its 17th

Biennial Conference in Montréal, 24–27 June 2008. The conference was

organized by the Université de Sherbrooke, on the initiative of Professor

Anastassios Gentzoglanis, Organizing Committee Chair. There was also a

main sponsor and corporate host – TELUS – represented by Dr Stephen

Schmidt, Organizing Committee Co- chair, and supported by Organizing

Committee Co- chair Stanford L. Levin, Southern Illinois University at

Edwardsville, and Program Chair Richard Schultz, McGill University.

Looking back, it was in 1997 that a major ITS conference took place in

North America. The industry and research have certainly moved a long

way, facing new technologies and regulatory challenges, but the critical

methods and policy analyses remain. The title of the conference aptly

refl ected both the change and the continuity: ‘The Changing Structure of

the Telecommunications Industry and the New Role of Regulation’.

The conference was a signifi cant success. The theme attracted a large

amount of submissions. In fact, it drew one of the largest pool of sub-

missions to an ITS conference ever, and the was one of the largest ITS

conferences ever from all over the world. Conference evaluations were

very positive after the event. The conference also brought to the fore new

research and policy agendas.

This book is based upon selected and edited conference papers. The

editors have been careful to select a wide range of chapters in order to

capture the convergence issue from a number of perspectives, and to

achieve a historical state of the art. The volume succeeds in providing a

multifaceted and rich view of both changing industry structure and for

new roles of regulation. The editors and the contributors are to be con-

gratulated for a timely book.

Erik Bohlin

ITS Chair

Professor, Chalmers University of Technology, Sweden

1

Evolving technologies, competition and the new role of regulation: introduction and synopsis of the book

Anastassios Gentzoglanis and Anders Henten

Regulation of the telecommunications industry has traditionally focused

on the supply side of the industry, chiefl y the retail segment of the market.

Since liberalization of the industry has begun, regulation has gradually

shifted to the wholesale segment of the market. The regulatory agencies

have intervened to regulate access and facilitate entry and, hopefully,

investment in infrastructure. This asymmetric regulation has had mixed

results. As an answer to that some regulatory agencies have abandoned

wholesale regulation (the case of the USA) and some others have shifted

from light- handed regulation to heavy- handed regulation with manda-

tory unbundling of the local loop (the case of Australia). Other regulatory

agencies have moved to a more gradual type of wholesale deregulation on

the grounds that competition in this segment of the market has not yet

fully developed, but as it grows the need for regulation is reduced (the case

of Europe and Canada).

Paradoxically, as competition increases in the retail segment of the

market, consumers are increasingly ‘impaired’ in their capacity to make

decisions in their best interest. The array of services and suppliers that

competition makes available increases consumer choice and with it the dif-

fi culties for them to make rational decisions with respect to services, quality

and prices, and so on. If consumers are able to make rational decisions,

competition among new service suppliers will increase. But if consumers

are unable to make rational decisions because of too much choice or poor

quality of information or misinformation, competition is dampened and

the competitive process is jeopardized. In that context, regulation must

emphasize the demand aspects of the industry and consumer protection

should become a priority for the regulatory agencies.

Further, next generation mobile networks (NGMNs) depend heavily

on the existence of ubiquitous broadband (BB) connectivity, applica-

tions and content. Broadband deployment is uneven within and among

2 Regulation and the evolution of the global telecoms industry

countries. Without an adequate deployment of BB technologies the eco-

nomic growth of the countries is jeopardized. Apparently, regulation is

one important factor determining the pace of deployment of BB technolo-

gies. The regulatory model chosen may impact positively or negatively on

investment in BB infrastructure. Additionally, ubiquitous BB and mobile

applications create new needs for spectrum availability and spectrum

management becomes an important function of the regulatory agencies,

particularly at this time of rapid evolution of the mobile technologies. The

NGMNs require a new distinctive regulatory and policy framework which

will deal explicitly with the issues and opportunities of the next phase of

wireless technologies.

Regulation evolves and its evolution is the result and the impetus of

change of the telecommunications industry structure and performance.

As the industry becomes more mature and incumbents and new entrants

get more familiar with the rules of the game, they become able to develop

strategies which increase the value of the fi rm. In a competitive context

telecommunications fi rms will invest only if their investment achieves an

average return which is greater than the weighted average cost of capital

(WACC). Projects with positive net present values (adjusted for specifi c

risks) are value accretive and therefore it is worth undertaking them. They

will bring more wealth to stakeholders through dividends and capital

gains. In a regulatory context the investment decisions, particularly for

projects of high risk (sunk investments in BB and NGMNs, for instance)

may not occur or may be ‘unreasonably’ delayed. Thus, under specifi c reg-

ulatory frameworks, deployment of BB technologies may not be optimal,

NGMNs may not roll out adequately or optimally, consumers may be

‘impaired’ or harmed; and these are not necessarily the results the regula-

tory authorities have sought to achieve in the fi rst place. Regulation thus

has a role in an evolving global telecommunications industry.

This was precisely the main theme of the International

Telecommunications Society (ITS) 17th Biennial Conference which was

held in Montreal from 24 to 28 June 2008. This book has been prepared

to highlight the main arguments and the richness of ideas about the new

issues which arise from the evolving structure of the global telecom-

munications industry and the role of regulation. The 15 papers from

the conference have been peer reviewed and they meet the stringent

criteria of the scientifi c papers and those of Edward Elgar Publishing.

We are grateful to the publisher and particularly to Alan Sturner for

his meticulous work in the preparation of this volume. We also thank

our colleagues and authors of the 15 chapters of this volume for having

contributed high- quality chapters. As organizer of the 17th Biennial

ITS Conference, Dr Gentzoglanis thanks TELUS, and particularly

Introduction and synopsis 3

Stephen Schmidt, for contributing fi nancially to the publication of this

book and its contribution, fi nancial and in kind, in being a corporate

host for the very successful ITS Conference. Gratitude is also extended

to the Université de Sherbrooke, Offi ce of Research and the Faculty of

Business Administration for their fi nancial contribution and support. Dr

Gentzoglanis thanks the SSHRC (the Social Sciences and Humanities

Research Council of Canada) for its generous contribution to the organi-

zation of the ITS Conference and the publication of this book. He also

thanks Emmanuella Gentzoglanis and Andrianiaina Rajaobelina for

their help and comments and suggestions. This book could not exist

without our many devoted manuscript peer reviewers. Sincere thanks to

all the reviewers who dedicated their time and shared their knowledge

with the authors and the editors of this book.

The fi rst chapter by Anastassios Gentzoglanis and Elias Aravantinos

provides an extensive review of the literature on the role of regulation in

the deployment of broadband technologies. The chapter starts out with

a general introduction to the role of regulation in telecommunications

markets and, furthermore, as an input discusses the relationship between

competition and investment. It is emphasized that this depends on the

degree of competition and the development of the market. An inverted

U- shaped curve is suggested as describing the implications of competition

on investment. More specifi cally, the chapter examines whether and how

broadband development can be promoted by means of regulation. Two

diff erent types of broadband competition are discussed, service- based

competition (SBC) and facilities- based competition (FBC). FBC is gen-

erally seen to be the most desirable situation or goal as FBC provides a

dynamic effi ciency while SBC is seen to lead to static effi ciency.

The chapter gives focus to the ongoing discussion as to whether the

static effi ciency associated with SBC can lead to the dynamic effi ciency

related to FBC. This is the proposition advanced by the theory of the

ladder of investment (LoI). According to this theory, new operators to the

market will not generally begin with investing in infrastructure but will be

more inclined to enter the market using a service- based model. Eventually,

such new operators will start investing in infrastructure when they acquire

a better understanding of the market conditions. Regulation can facilitate

this process by setting the right entry conditions, for example, by start-

ing with a relatively low access fee which is later on increased as the new

entrants settle in the market.

The chapter examines the theoretical aspects of the ladder of investment

theory and also refers to papers presenting empirically based analyses of

the tenability of the theory. Moreover, the chapter introduces the issue

of intermodal or interplatform competition. Even though this issue is not


absent per se from the LoI theory, nonetheless, when debating LoI, there

is a tendency to confi ne it within a single type of access network. The

chapter by Gentzoglanis and Aravantinos contributes to the debate by

synthesizing the main arguments pertaining to the LoI theory. In addition,

the chapter underlines the fact that BB development in diff erent countries

is not the same and this is attributed to the diff erences in the countries’

regulatory policies and the type of their initial technological infrastruc-

ture. It should be expected therefore that the implications of a LoI policy

will be diff erent for countries with basically just one type of infrastructure

(PSTN – public switched telephone network) and countries with compet-

ing infrastructures (for example, PSTN and cable).

On the basis of the theoretical propositions and the empirical analyses

referred to, the chapter concludes that the actual LoI policies implemented

so far have not been successful, but one cannot dismiss the possibility

that a LoI policy can work. This presupposes, however, that the timing is

right and the policy is fi ne- tuned precisely with respect to the types of new

entrants. Regulators should then ask: Are they fi rst- movers or second-

movers? But this sets very high requirements on regulation.

In the second chapter, William Taylor discusses the nature of intermo-

dal competition and its implications for regulation of wholesale services in

the context of the US, but given the ubiquity of intermodal technologies,

the conclusions of his analysis can fi nd a wider application. Taylor notes

that demand for wholesale services is a derived demand, in the sense that

wholesale services can be considered ‘essential’ in the provision of retail

services in the downstream market. Competition among retail service

providers may curtail retail prices but there is a likelihood of anticom-

petitive conduct when the incumbent local exchange carrier (ILEC) is not

regulated in the form of mandatory unbundling and provision of essential

wholesale services at regulated prices.

When competition is fi erce among dependent wireline competitive local

exchange carriers (CLECs), their returns in the retail market cannot be

higher than normal and therefore the ILEC monopolist of wholesale

wireline services cannot behave anticompetitively and charge high prices

for its essential wholesale services. Since it cannot exercise its market

power at the wholesale level, its profi ts will also be normal. This may

have some undesirable eff ects on investment in infrastructure. The man-

datory unbundling of ILEC facilities at regulated rates would reduce the

incentive of retail wireline competitors to invest in their own network

infrastructures and compete on an end- to- end basis. At the same time, the

requirement that ILEC facilities be shared with competitors reduces the

ILEC’s incentives to introduce and roll out that infrastructure, particu-

larly for services associated with investment that will, eventually, be sunk.


Further, Taylor casts doubts about the necessity of price regulation at

multiple stages of production (wholesale and retail). Using examples from

the US, he demonstrates that regulation at the wholesale and retail levels

becomes a source of unintended consequences in competitive markets.

Thus, where intermodal competition is present the ILEC’s capacity to

exercise its market power is nil and therefore customers cannot be harmed.

Accordingly, ex ante economic regulation cannot lead to an increase in

social welfare and there are not any effi ciency arguments that can justify

the presence of economic regulation at the wholesale level. In the absence

of an ex ante wholesale regulation, customers can be protected from

competition authorities which will respond ex post to specifi c complaints

of anticompetitive behavior on behalf of ILEC. Thus, where intermodal

competition exists, ex post regulation is a better vehicle to protect custom-

ers than ex ante regulation.

In the third chapter Martyn Taylor delves into the question of how to

strike a balance between competition, investment in infrastructure and reg-

ulation. Using the conventional fi nancial framework for project appraisal,

Taylor argues that the returns that a fi rm expects to get from investment in

infrastructure are constrained by market conditions (consumers’ willing-

ness to pay), the presence of substitutable services (wireless broadband,

for instance) and regulatory policies. These factors increase the perceived

risk of the investment in next generation networks (NGNs) and the latter

will not materialize unless a fi rm has reasonable certainty that it will earn

a return that exceeds its cost of capital over the project’s life.

Taylor argues that well- intended regulation can have unintended adverse

eff ects, chiefl y by providing disincentives to investment due to an increase

in perceived risks particularly when regulation is asymmetric. Network

unbundling and cost- based access regulation are two cases in point. While

they contribute to allocative (static) effi ciency by increasing the level of

competition to which incumbent network owners are subject, they may

do so at the expense of long- term investment incentives. The total element

long- run incremental cost (TELRIC) approach has been used in the US

and its variant – total service long- run incremental cost (TSLRIC) – in

Australia as a mechanism to regulate prices based on the incremental

costs faced by an effi cient cost- minimizing fi rm with an optimally con-

fi gured network that uses the best available current technology. Under

the TELRIC approach, any investment in network infrastructure would

simply earn enough to cover the project’s weighted average cost of capital

(WACC), bringing thereby the net present value (NPV) of the project

near zero. The fi rm would not have an incentive to invest in infrastructure

and the desirable long- term eff ects of dynamic competition are sacrifi ced.

This was the case with Australia which has applied excessive regulation


using the TSLRIC approach. The application of TSLRIC pricing by the

Australian Competition and Consumer Commission (ACCC) has deterred

investment in infrastructure and investment fl ows have been distorted

away from regulated (and potentially regulated) services towards unregu-

lated services and infrastructure. The negative eff ects of excessive regula-

tion are illustrated by the example of failed negotiations between Telstra

(Australia’s incumbent telecommunications carrier) and the Australian

government when the former was seeking exceptions from the current

regulation in order to implement fi ber to the node (FTTN) upgrade

investments. Telstra’s withdrawal forced the government to intervene and

proceed with the investment in partnerships with the private sector. Taylor

suggests that policies such as access holidays and public–private partner-

ships (PPPs) may be some interesting techniques to provide incentives to

fi rms to invest in NGN infrastructure.

Patrick Xavier and Dimitri Ypsilanti’s chapter (Chapter 4) examines the

demand side of the telecommunications industry. Since liberalization, the

number of new entrants in telecommunications markets has increased con-

siderably and service competition has grown signifi cantly. In this context,

Xavier and Ypsilanti believe that the consumers’ rational decision- making

mechanism is seriously impaired amidst the plethora of services and pack-

ages off ered by the telecommunications fi rms. This contrasts with the

neoclassical view of consumer behavior according to which consumers

are able to make rational decisions and choose, all the time, the goods

and services which maximize their utility. In reality, consumers’ behavior

departs signifi cantly from the traditional rational behavior assumed by

the conventional neoclassical economic theory. By using the behavioral

approach to consumer behavior and statistical examples, the authors dem-

onstrate that consumers choose telecommunications services for a number

of reasons which may not be characterized as a ‘rational’ behavior in the

sense of classical theory. For instance, consumers are discouraged from

switching to a diff erent service provider because of the perceived or real

high switching costs (lengthy and cumbersome switching procedures; early

exit charges; confusing products and non- transparent pricing; technical

incompatibility of equipment; long- term deals that lock consumers into

lengthy relationships with their providers).

The barriers to switching service providers can also have an impact on

the supply side of the industry. New entrants could be deterred from enter-

ing the market, fearing that they will be unable to persuade customers to

switch from their existing provider. ‘This could diminish contestability

and the eff ectiveness of competition and limit the benefi ts that consumers

would otherwise derive from it’, say Xavier and Ypsilanti. The statistics

from various countries (the UK, Portugal, Australia and the US) and


from various segments of the market (fi xed line, mobile and Internet)

illustrate that competition has made the telecommunications industry

quite complex (on the demand side) and consumers facing this complexity

prefer to adopt consumption strategies which are not necessarily in their

best interests. Indeed, consumers will prefer to stay with what they know

(no switching) instead of choosing a cheaper alternative.

Xavier and Ypsilanti’s analysis is consistent with the argument of

behavioral economics according to which an ‘endowment factor’ serves

to infl uence decisions in favor of the present provider. In that context,

regulation has an important role. Regulators have to take a number of

measures to assist consumers to make ‘rational’ decisions in their best

interests. Educating consumers; increasing awareness about new services

and options; requiring that all major operators provide comparable and

complete information about the services, quality and prices; targeting

information to most vulnerable consumers; could be some of the new

responsibilities of the regulatory agencies in an era of increased competi-

tion in the telecommunications industry.

In the fi fth chapter Marcel Boyer develops a methodological frame-

work which is used to characterize properly the level of competition in the

telecommunications industry, particularly in the residential local access

market. He argues that failing to recognize and properly evaluate the

nature of competition in the telecommunications industry may lead to an

ineffi cient use of regulation. In competitive telecommunications markets,

regulators play a new role and must undertake three new functions, acting

as: generators of information for the consumers; managers of the rules of

competition among telecommunication players; and promoters of effi cient

investment programs.

Boyer argues that the telecommunications industry has changed dra-

matically from the mid-1990s onwards and the technological changes

have made it appear much more like an emerging industry than a mature

industry. Therefore, the traditional measures used to determine the level of

completion in the industry are less relevant. New measures must be used to

determine the level of competition in the industry. For instance, the use of

market shares as an index of competition may make sense in mature indus-

tries where there is a relative stability of market conditions, but applying

it to the telecommunications industry characterized by a rapid pace of

changes may be misleading. Further, even though the pricing schemes

used by competitors are diff erentiated so that switching among service

providers becomes diffi cult on the part of consumers, the lack of switch-

ing does not make the telecommunications markets less competitive. On

the contrary, the very existence of price diff erentiation indicates that the

telecommunications industry is indeed competitive. Without it, prices


would have been quite low, making the provision of new services a losing

opportunity inviting exit from the industry and an ultimate reduction in

the number of fi rms in the industry (less competition).

Although competition may be limited in terms of prices, telecommuni-

cation service providers may compete in a number of other areas, such as

coverage, type of transmission (digital or analogue); interplatform provi-

sion of services (DSL – digital subscriber line; cable modem), security,

and so on. Therefore, traditional measures of measuring competition such

as relevant market, relevant (substitutable) services available to consum-

ers, relevant set of actual or even potential competitors, and so on, are less

relevant during this changing phase of the telecommunications industry.

In that context, Boyer proposes a diff erent regulatory mechanism. He

argues that, to achieve a proper balance between static (short- term) and

dynamic (long- term) goals, regulators must rely on competitive proc-

esses, in a sense that instead of micro- managing prices and quantities,

they must make sure that these prices and quantities emerge from a

competitive environment. Thus regulators have an essential role: they

have to act as effi cient generators of social effi ciency by safeguarding the

competitive process in the telecommunications industry. This is achieved

by making sure that inter- access to essential facilities is available at non-

discriminatory conditions and prices so that only new, more effi cient

entrants enter the industry. In that way, consumers benefi t from the entry

of effi cient competitors. Thus, viewing the emerging structure of the tel-

ecommunications industry as a process, regulators must act as: (1) trusted

generators of information for consumers; (2) managers of fair conditions

for access to the local loop; and (3) promoters of investment programs

which should contain pricing rules designed to include all network access

costs and guarantees safeguarding the integrity and reliability of the entire

network.

The sixth chapter, contributed by Kenneth Jull and Stephen Schmidt,

intends to off er an alternative framework that can be used as a basis for

new approaches to the regulation of the telecommunications industry. The

authors make a distinction between ex ante and ex post regulation and

they use examples from other industries to illustrate that ex post regulation

can be an option for the telecommunications industry. Traditionally, the

regulatory systems in the telecommunications industry have been ex ante

systems, but the latter are increasingly criticized as ineffi cient since they are

perceived as being a blunt ‘one- size- fi ts- all’ mechanism. In Canada, there

are suggestions (Telecommunications Policy Review Panel) to replace this

mechanism by a new regulatory framework which would set out broad

principles to prohibit anticompetitive conduct instead of detailed ex ante

rules.


Jull and Schmidt propose three principles that ought to govern the

balance between ex ante and ex post systems in the application of the regu-

latory policies. According to them, regulators focus ought to be on: (1) the

prevention of harm and attainment of specifi c social objectives; (2) strate-

gies for managing risks; and (3) the adoption of fl exible mechanisms (use

of multiple models) to refl ect the diff erent needs of the stakeholders.

Each system, being ex ante or ex post, might have two subsets within it,

being rules- based and principles- based. A rules- based system (whether it

is ex ante or ex post) is better suited to industries which are stable, techno-

logically simple and where the economic and fi nancial stakes are relatively

low. The telecommunications industry is undergoing rapid technological

and market structure changes, so this industry is neither simple nor stable

and the economic stakes are very high. Because rules are infl exible, they

can be overtaken by changing circumstances in fi elds such as telecom-

munications. In such a context, a principles- based approach may be more

appropriate. As a matter of fact, the choice between principles- based and

rules- based systems is a function of social and economic priorities and

the level of maturity of the industry. Rules- based systems are more suit-

able for cases of social regulation where serious harm may be prevented

whereas principles- based systems are generally more suitable for cases of

economic regulation. Jull and Schmidt propose the same type of regulation

for the Canadian telecommunications industry as the one proposed by

Boyer in Chapter 5 but each author draws his conclusions using a diff erent

analytical framework.

The seventh chapter by Olaf Rieck analyzes, from the industry’s per-

spective, the strategic activities of the telecommunication carriers to

integrate vertically the various segments of the value chain. Few studies

examine quantitatively the new strategic directions that telecommunica-

tion fi rms take in the rapid changing environment, and Rieck’s study

falls into this category. He divides the telecommunications industry into

fi ve layers and then uses Fransman’s (2007) value chain simplifi ed layer

model to assess empirically the impact of various strategic initiatives on

the valuation of telecom operators. Rieck is interested in the evolution of

the structure of the telecommunications industry and in the strategies of

industry players to extend their market power to more layers. This is quite

interesting when one views the recent changes in the telecommunications

industry structure and its trend to convergence.

Nowadays, the telecommunications industry has many fi rms which

entered from outside the traditional telecommunications services industry,

and many are quite new or did not even exist back in 1998. For instance,

Google emerged as world leader for online searches and has established

itself as a signifi cant player in the telecoms value chain. Google is also


involved in the rollout of broadband infrastructure and the provision of

content. Apple launched the iPhone which threatens various traditional

players in the value chain. By off ering hand phone devices, Apple has

eff ectively become a new player in the hand phone equipment market

(Layer 1). By tying Apple’s handsets to iTunes (Layer 4) and by strik-

ing content deals with content providers (Layer 5), Apple has extended

its market power in almost all layers of the value chain. Nokia and Sony

Ericsson have followed Apple’s lead by launching their own content plat-

forms. Traditional telecommunications carriers, while under threat from

all sides to be reduced to ‘bit- pipes’, have tried to counter the threats by

engaging in activities in vertically related markets. This includes initiatives

like joining the open handset alliance, the development of mobile portals,

or striking deals with content providers such as to strengthen their posi-

tion in the content integration layer. Do these strategies help to increase

the telecommunications fi rms’ reach and market power, or will they reduce

their role in the future? Rieck answers these questions by examining the

reactions of fi nancial markets to the strategies adopted by the telecom-

munications companies, particularly Mergers and Acquisitions (M&A),

after rival fi rms have announced their decisions to extend their reach in the

vertical value chain of the industry.

Chapter 8, authored by Bronwyn Howell, asks the question whether an

industry- specifi c regulatory regime (a telecommunications regulator) is

more able to pursue an economic effi ciency (static and dynamic) objective

than a competition authority (non industry- specifi c) without falling into

the trap of regulatory capture. To answer this question, Howell examines

New Zealand’s telecommunications sector in the 1990s and 2000s. She

notes that the initial goal of New Zealand’s government was to preserve

the telecommunications industry’s long- run incentives to invest in new

networks and technologies using a ‘light- handed’ regulatory regime.

Unfortunately, the objectives of regulators often do not coincide with

those of politicians. The latter are subject to more pressures from vested

interest groups and more inclined to satisfy their demands by adopting

new and/or modifying existing legislation to pursue diff erent sectoral

objectives. Given that regulators are the agents of political principals, they

may lose their power when politicians decide to change objectives and

move from effi ciency objetives to distributional ones, as was the case with

New Zealand’s government in the 2000s. Indeed, prior to the competition

law review in 2000, New Zealand had adopted ‘light- handed’ regulation

for its telecommunications sector. Despite this regulation, the telecommu-

nications industry was far from unregulated. Under a contractual arrange-

ment, the incumbent monopolist, Telecom Corporation of New Zealand

Limited, had rural–urban universal service obligations, free local calling


and a price cap on residential services. This contractual arrangement,

known as the ‘Kiwi Share’, was a type of regulation capable of achieving

economic effi ciency.

The objectives of regulation have been changed with the change of gov-

ernment in 2000 and the adoption of the Telecommunications Act in 2001.

The Act established an apparently independent Telecommunications

Commission (within the Commerce Commission) and TSLRIC pricing

for ‘designated services’. This creation of an apparent independent regu-

latory body free from risk of capture by vested interests was seen as an

‘enlightened’ form of industry- specifi c regulation. Nonetheless, its inde-

pendence was tested when the Commission undertook a revision of local

loop unbundling (LLU). The Commission, applying dynamic effi ciency

principles, decided not to proceed with unbundling. By contrast, the

Commission used a diff erent approach when it had to examine the mobile

termination market where serious concerns have been raised concerning

the exercise of monopoly power. In a surprising decision, the Commission

asserted that the sector’s objective was to pursue competition rather than

effi ciency, and therefore short- run objectives were prioritized. Although

making effi ciency an explicit regulatory objective is rationally justifi ed

from an economic perspective, viewed from a political perspective the

effi ciency objective is unsustainable in the long run. The New Zealand case

clearly illustrates that ex ante regulation can be inferior compared to ex

post regulation – competition law – particularly when the risks of regula-

tory capture cannot be avoided.

Scott Marcus and Dieter Elixmann contribute the ninth chapter. They

argue that the migration of current networks to next generation networks

(NGNs) and the issues arising from their access to the fi xed network bring

new challenges to regulators and policy makers alike. For instance, the use

of local loop unbundling (LLU) as a solution to the problem of incum-

bent’s market power is particularly challenged by the migration to FTTC/

VDSL (fi ber to the curb/cabinet/very high speed digital subscriber line)

or to FTTH/FTTB (fi ber to the home or fi ber to the building) networks.

The migration to NGNs is very diff erent from country to country and

these diff erences are due to the existence of various regulatory regimes and

the market evolution in each country. Germany and the Netherlands, for

instance, have a regulatory framework which incentivizes incumbents to

be the ‘fi rst- movers’ to replace the traditional fi xed access network with

VDSL- capable networks. In France, the NGNs’ deployment has been

undertaken by the incumbent as well as its competitors. In Japan, deploy-

ment is realized not by the incumbents but by other independent compa-

nies. In the US, a change in regulation in 2003 put an end to mandatory

broadband unbundling and allowed broadband services to be off ered over


cable. Such a policy gave incentives to incumbent operators AT&T and

Verizon to make substantial investments in fi ber. The authors compare

the performance of each country trying to identify whether competition

and/or regulation are the most important factors for change. After having

analyzed in detail the level of deployment of the fi ber- based NGNs and

the regulatory regimes determining the access conditions in Germany,

the Netherlands, France, Japan and the United States, they conclude that

regulation is an important input in the deployment process of NGNs but

there is no unique model of regulation which fi ts well to every country.

Rather, a number of factors – demographic, geographic and historical

(availability of alternative last mile infrastructure) – determine the for-

mulation of the regulatory policy. Once the latter is well conceived and

put into play, it is the force of competition which determines the pace of

NGNs deployment. Thus, country- specifi c regulation and competition are

the sine qua non conditions for a wider deployment of NGNs.

The tenth chapter is by Arata Kamino and Hidenori Fuke, and is the

fi rst of three chapters to discuss functional and structural separation and

its eff ects on the rollout of BB technologies and platforms and NGN

applications. Their analysis is a case study of the Japanese telecommunica-

tions industry demonstrating that the deployment of BB technologies and

the implementation of the world’s fastest and cheapest DSL technologies

in Japan are attributed to the particularities of the Japanese regulatory

regime and market conditions. Indeed, Japan was one of the fi rst indus-

trialized countries to implement the most rigid open network policy for

the promotion of service- based competition. For instance, unbundling

obligations have been imposed on both copper and fi ber loops and the

competition that this regulation has entailed resulted in very low LLU fees

particularly for shared lines. Given the rapid increase in competition in

LLU, no bitstream access competition has been developed in Japan. The

European Union (EU) and the US have implemented a similar type of

copper LLU but they have not experienced the same degree of DSL as in

Japan. Although there is no simple answer to this conundrum, the authors

advance the arguments according to which the diff erence in performance

may be attributed to the way this regulation has been applied in diff er-

ent countries and continents. For instance, in Japan, the LLU fees were

fi xed at an extremely low level favoring service- based competition by new

entrants. In addition, the Japanese entrepreneurs may be driven more by

a kind of ‘animal spirit’ which contributes to intensify competition even

when profi ts are not as high as they could be. Indeed, Japanese DSL com-

petitors continued to provide very cheap alternative DSL services despite

long- standing fi scal losses during the early 2000s. Competition is eff ec-

tive and works when it is fair and transparent. But it is not entirely clear


whether competition is more eff ective when conduct regulation is imposed

on the incumbent or when structural separation is applied. Given that

Japan adopted in 1999 the same level of conduct regulation as the EU, but

that competition and the deployment of BB technologies have developed

faster in Japan than the EU, the authors argue that Japan’s structural

separation was an eff ective means to foster competition in the market.

Nonetheless, the introduction of FTTx (fi ber to the x) and NGNs raises

additional issues in the discussion of vertical separation in the telecom-

munications industry, such as the ‘hold- up’ and ‘coordination’ problems,

and one should analyze the perspectives of competition which will emerge

between traditional carriers and content providers on the basis of new

business models before the implementation of a structural separation.

Peter Curwen and Jason Whalley continue the theme of the implemen-

tation of functional separation in Chapter 11, this time using the UK

experience. In the EU, the implementation of functional separation within

fi xed telecommunications markets is increasingly seen as a way to resolve

the tensions that exist between incumbent operators and those other

service providers that require access to incumbents’ networks to deliver

their own services. In 2005, functional separation was implemented by

the British incumbent, British Telecom (BT), after pressure by Ofcom,

the regulatory authority, thereby making the UK the de facto European

leader in functional separation. Under this agreement, BT created a new

company – Openreach – to run BT’s local access network. The creation

of Openreach was possible after BT has agreed on a series of undertak-

ings. Curwen and Whalley’s chapter focuses on the implementation of

these undertakings and highlights the diffi culties encountered to make the

functional separation eff ective. The authors demonstrate that functional

separation is not a simple task. On the one hand, regulators encountered

enormous diffi culties in making the undertakings operational and, on the

other hand, Openreach had diffi culties in implementing them.

One of the major tasks in implementing functional separation is to

selectively separate those parts of the network that are diffi cult for other

operators to replicate but which they need to access in order to provide

their own services. To interpreter such a separation as simply a division

of the incumbent’s wholesale and retail businesses from one another –

either in the form of accounting, corporate or type of service (local from

long distance; mobile from fi xed; local from broadband, and so on.) – is

completely misleading. The way functional separation is defi ned – in a

broad or narrow sense – has important implications for the actual form

of separation. Regardless of the extent to which functional separation

is implemented, regulators should provide incentives so that the sepa-

rated network could act in the interests of all its customers, internal and


external, and not in the interest of its parent company. BT’s separation

was driven by the need to incorporate EU directives into the UK regula-

tory framework, the failure of competition to develop as anticipated in the

UK, and the establishment of Ofcom in 2003, which undertook a strategic

review of the telecommunications market in order to examine the level of

competition of the telecommunications industry in the UK and identify

the regulatory options available. Ofcom concluded that deregulation was

not possible because sector- specifi c regulation was faster and more precise

than the alternatives – competition. Ofcom opted for a functional separa-

tion called ‘real equality of access’ under which independent purchasers of

BT’s wholesale products could buy these products under the same terms

as BT’s own retail operations. In practice, the ‘real equality of access’ took

the form of equivalence of outcome and equivalence of input. Under the

former, wholesale customers receive products that are comparable to those

off ered to BT’s own retail operations, but the underlying processes are dif-

ferent. Under the latter, wholesale customers receive the same products

as BT’s own retail operations using the same set of underlying processes.

Despite the delays and the defi nition problems, the period subsequent to

the adoption of functional separation has seen the emergence of signifi cant

broadband competitors. LLU played a central role in the strategies of

entrants and off ered incentives to other service providers to invest in other

parts of the ‘ladder of investment’. The emergence of LLU as a vehicle for

the deployment of broadband services lifted the importance of BT in the

market and made sure that the relationship between BT and those compa-

nies using its network was functioning as planned.

Toshiya Jitsuzumi is the author of the twelfth chapter. This chapter

provides a theoretical justifi cation and an econometric analysis of the

hypothesis that LLU is contributing to the deployment of new technolo-

gies such as FFTx. He emphasizes the need to take into account social and

economic aspects of the issues arising from the network neutrality debate,

and he concentrates his analysis on the solutions that are most effi cient

from an economic point of view in the short term and long term. In the

short term, it is assumed that entry does not occur and the market is served

by the existing network operators. In the long term, entry occurs, competi-

tion is more intense and network congestion becomes a less acute issue.

The proposed solutions are dependent on the assumptions made.

For instance, when the network capacity is fi xed in the short run, the

problem becomes one of static effi ciency maximization. In that case

network operators have market power and regulators have to fi nd effi -

cient solutions to discipline incumbents who control bottleneck facilities.

But given that content providers depend on the presence of other fi rms in

the industry in the provision of their own content, it is unlikely that they


will exercise their market power and foreclose the market from competi-

tion. Indeed, in this industry, a fi rm is interested in the internalization of

complementary effi ciencies (ICE) arising from applications created by

others. This behavioral characteristic – the ICE in conjunction with the

costs associated with regulation and information asymmetry – makes the

competition solution a more desirable solution than regulation. Knowing

that a profi t- seeking bottleneck monopolist acts to maximize its effi ciency,

regulators should not use rules- based regulation but rather principles-

based, unless there is pressure arising from the exercise of signifi cant

market power (SMP) by some service providers. ICE makes regulation less

desirable and government intervention is kept to a minimum.

In the long run, the issues become more challenging than in the short

run. In the long run, market conditions must be such that incumbents and

new entrants have the appropriate incentives to invest in new technologies

and maximize dynamic effi ciency. Investments, being in the form of virtual

capacity, better protocols – peer to peer (P2P), proactive network provider

participation for P2P (P4P) – and/or better network management, must

be fi nanced but the problem is that, as yet, there is no any sound business

model that provides incentives to stakeholders to invest in infrastructure.

Network neutrality proponents suggest the use of a subscription model –

that is, through additional monthly subscription revenues such as quality-

of- service (QoS) surcharge from end- users – while the opponents suggest a

business model according to which investment in infrastructure is fi nanced

by charging content and application providers. Neither of these models

can assure the collection of suffi cient revenues for capacity expansion or

the quality of the transmission of the content. These solutions may be

utility- decreasing and unsustainable in the long term.

In order to verify whether the subscription model is sustainable in the

long run, Jitsuzumi conducted a survey of Japanese broadband users using

an e- mail and web- based system. His data and econometric analysis indi-

cate that such a business model is not sustainable in the long run unless

there are positive expectations concerning the future technological devel-

opments and the existence of a fund allocation mechanism. His results are

quite informative and useful for other countries.

Bruno Basalisco, Andy Reid and Paul Richards contribute the thir-

teenth chapter, delving into the interesting question about the eff ects of

regulation on innovation and on evolving technologies, most of which

emerge outside the domain of the telecommunications industry. Departing

from the self- evident fact that the main objective of innovators is the com-

mercialization of their innovations, they argue that such commercializa-

tion is more successful when various parties across sectors coordinate

their activities. The converging nature of new technologies requires a keen


interplay between innovation and regulation and makes the latter less

desirable, particularly when the boundaries of the industries are merging.

In an increasingly competitive world, erstwhile competitors realize that

innovation requires more than large market shares and market domi-

nance, it needs ‘co- opetition’. The latter is used to create a competitive

advantage in innovation processes involving all ‘co- opetitors’, that is, their

network of suppliers, users and customers. Collaborative ventures with

other industry stakeholders lead to greater coordination of the innovative

activities and reduce the uncertainty associated with technological changes

across sectors. Innovation interdependence delivers maximum benefi ts

to their participants. The standardization of technologies through col-

laborative ventures and interconnectivity minimize the level of risk of the

entire value chain and not just one part of it. But regulatory decisions may

contribute negatively to the tendency of collaborative innovative agree-

ments and thwart the appearance of new business models. The presence

of network spillover eff ects in the various processes of innovation provide

incentives for diff erent networks to interconnect. Further, the presence of

path- dependence in innovations implies that successful innovation by one

or more player(s) is likely to infl uence future technological choice across

the industry. Co- opetition, path- dependence, spillover eff ects and stand-

ardization add considerable complexity to the exploitation of network-

based innovations, but reduce the risks associated with innovations in

converging industries. Basalisco, Reid and Richards argue that the regula-

tory frameworks should be operated in such a manner that diff erent risks

associated with diff erent innovations are treated diff erently in regulatory

terms. Failing to recognize this means the current regulation may provide

disincentives to fi rms to commit resources to invest in innovative activities

and infrastructure.

The penultimate, fourteenth, chapter by Claudio Feijóo, Sergio Ramos

and José- Luis Gómez- Barroso adopts an interesting approach to examine

the impact of regulation on the pace of deployment of the next generation

mobile networks (NGMNs). They argue that regulators may adopt either

a stable and coherent framework which provides incentives to invest-

ment in broadband technologies, or a framework which may retard the

deployment of these new technologies by amplifying all techno- economic

uncertainties. According to the authors, spectrum management is one

of the most important areas where regulation may have an immediate

impact. Referring to the regulation of spectrum in the EU, the authors

urge regulators to make spectrum management more fl exible in order to be

able to accommodate the next generation mobile networks (NGMNs) and

other fast- evolving technologies. They also suggest a better harmonization

of spectrum management mechanisms across member states for easier


deployment of ubiquitous broadband infrastructure and faster realiza-

tion of the benefi ts of the NGMNs. Drawing from the US experience and

the relative failure of the Universal Mobile Telecommunications System

(UMTS) as opposed to the great success of Global System for Mobile

Communications (GSM) in Europe, the authors argue that harmoniza-

tion should be subtler and focus mostly on the new conditions for use of

spectrum and particularly on the ‘converging competition’ which emerges

from the fi xed–mobile convergence of NGNs. Uncertainties caused by an

ill- conceived regulatory framework which does not take into account the

conditions for investment in NGMNs, or the conditions for competition

and the conditions for innovation, will retard the NGMN deployment

and it will result in a loss of the EU’s competitiveness. Regulation has an

important role to play in the intensity and speed of the NGMN deploy-

ment and can contribute signifi cantly to the creation of value through the

arrival of new applications and services. To do so, regulation must be swift

and adaptive to the requirements the new technologies bring about.

In the fi nal chapter, Chapter 15, Morten Falch, Anders Henten and

Karsten Vandrup address the deployment of mobile data in Europe, East

Asia and North America in order to identify reasons that may explain

the diff erence in performance among these regions. The chapter focuses

on the market conditions that prevail in each geographical market but

it also addresses the role of policy in the deployment of data and mobile

Internet. It is argued that to promote mobile data one needs to encom-

pass a wide variety of policy areas since the development of mobile data

depends on structural factors in the markets, particularly those pertaining

to the supply side of the industry and the take- up factors like the general

e- readiness of the potential users on the demand side.

The chapter notices and documents that the East Asian countries,

Japan and Korea, are ahead in the global development of mobile data.

While Europe took the lead in the 1990s with the second generation GSM

system, the East Asian countries have been the frontrunners with respect

to mobile data, fi rst on 2.5G platforms and later on 3G and 3.5G plat-

forms. The US has been trailing somewhat behind in the 2G development

but seems to catch up regarding mobile data. The question is what expla-

nations there are for this development and what one can learn from it.

In order to examine this question, the chapter fi rst discusses the reasons

for the East Asian lead as compared to Europe and North America. The

chapter includes explanations of a structural kind on the supply side as

well as diff usion issues on the demand side. Theoretically, the chapter

therefore takes its points of departure in theory on innovation systems

and theory on diff usion. Secondly, the chapter examines empirically the

development of mobile communications more generally, and mobile data


more specifi cally, in the East Asian countries, Europe and North America.

Focus is on the terminal markets, network infrastructures, and services

and content. The numbers show that the US is forging ahead regarding

mobile data and that the European lead over the US in mobile communi-

cations is disappearing.

The prime reason put forward in the chapter for the North American

catch- up is the position of the US in the information technology (IT) area,

hardware as well as software. The US is positioned very strongly regard-

ing Internet technologies and services, and the hypothesis is that the US

can leverage this position onto the mobile fi eld. The Internet innovation

system and the mobile communications innovation system have to a large

extent been separate. The East Asian countries have managed to merge

them to a certain degree. However, the competences on the supply as well

as the demand side in the US with respect to Internet technologies consti-

tute a strong point of departure for developing mobile Internet in the US.

In sum, Regulation and the Evolution of the Global Telecommunications

Industry is a collection of 15 chapters that bring a variety of theoretical

perspectives and empirical evidence to the question of how regulation

could be applied (even eliminated) to the deployment of BB technolo-

gies and NGMNs in an era of dramatic changes in the structure and

performance of the global telecommunications industry. The material

well illustrates the diversity of thoughts and research that characterize

this important area of academic and business research. We hope that this

volume will spur others on to research this challenging topic.

PART I

Regulation versus investment: the balance between static and dynamic effi ciencies and the main issues of regulatory policy

21

1. Investment in broadband technologies and the role of regulation1

Anastassios Gentzoglanis and Elias Aravantinos

1.1 INTRODUCTION

In the two decades from the mid- 1980s to the mid- 2000s network indus-

tries have undergone dramatic changes in structure and regulation. The

thrusting changes have initially occurred in the technologies used by

these industries to produce, transmit and distribute essential services to

consumers. Technological changes have not uniformly aff ected all seg-

ments of the production processes. Few segments still remain natural

monopolies while others are more suitable for competition. This uneven

impact of technological changes on the cost functions has created new

challenges to regulators and the industry. Both still strive to fi nd a ‘new’

business model and new regulatory frameworks which will allow more

investment in infrastructure in general and in broadband (BB) technolo-

gies in particular.

Some countries have applied ‘light- handed’ regulation while some

others have even decided to go further and liberalize entirely a few seg-

ments of the telecommunications industry by introducing full competi-

tion while keeping others under a regulated monopoly regime. The latter

was generally applied to the incumbent while new entrants enjoyed more

favorable entry and access regimes. The duality created by this type of

regulation has sparked a heated debate as to what is the ‘appropriate

role’ of regulation and its impact on investment and innovation.2 The

main issues concern the capacity of the regulatory agencies to promote

investment in infrastructure, especially in broadband technologies, and

what are the best mechanisms to increase consumer choice at reason-

able prices. This debate has arisen because more and more specialists

(Aron and Crandall, 2008) believe that the telecommunications industry

has entered into a maturity phase, where competition can work really

well. In that case, innovation and dynamic effi ciencies are viewed as the


outcome of competition. Yet, there are other specialists who believe

that dynamic effi ciencies can only be achieved through an active use

of an incentive regulation (Cave, 2006). In the fi rst case, competition

spurs the diff usion of innovations and guarantees the best outcome for

customers. In the second case, discriminatory regulation (the one favor-

ing new entrants at the expense of the incumbent) is considered as the

most appropriate means to promote BB investments and effi cient prices

to customers.

But the introduction of competition modifi es the vested interests of

stakeholders, and regulators have to weigh the interests of entrants,

incumbents and customers in the design of their new regulatory frame-

work. If the new regulatory regime encourages entry by allowing access

at low prices, competition would increase, but the latter acts as a deter-

rent to investment in infrastructure since low access prices reduce the net

present value (NPV) of the incumbent’s investment projects. As a conse-

quence, the conditions to achieve competition and encourage investment

in BB are not very clear for the regulatory agencies, and their favored

policy to create equal access conditions to new entrants by unbundling

incumbent’s infrastructure has become a challenging task for regulators.

It is true that in many instances, wireline access is still the key infra-

structure to provide BB services.3 This bottleneck infrastructure is usually

controlled by the incumbent and confers on it an undue market power. By

regulating the access price, regulators aim at providing incentives to new

entrants to use the current technological platform and create competition

at the local level, but this may have a negative eff ect on infrastructure

investment. Investments in infrastructure and the diff usion of broadband

technologies and services have important externalities and spillover eff ects

and both are increasingly viewed as important factors for economic growth

and prosperity (Waverman et al., 2005). It is not surprising therefore to

fi nd out that the regulatory systems that allegedly impede investments are

under great scrutiny. This debate boils down to questioning the effi ciency

of regulatory policies and whether they are still justifi ed under the current

state of diff erent technological platforms. It is unclear whether they con-

tribute to or inhibit the development of innovation and infrastructure in

BB technologies and whether competition can better attain these dynamic

effi ciencies in the telecommunication industry.

The approaches adopted by the regulatory agencies in Europe and the

US to promote the deployment of BB technologies are quite diff erent.

In Europe, where traditionally digital subscriber line (DSL) technology

was most prominently used, service competition seems to work relatively

well and the rate of deployment of BB technologies is considered quite

satisfactory, compared to other countries. In North America, service

Investment in broadband and the role of regulation 23

competition has not given the anticipated results and at the time of

writing North America trails Europe and Asia. Therefore, the North

American regulatory agencies increasingly substitute service competition

for facilities- based competition. It is believed that facilities- based compe-

tition will attract new entrants whose investments in new infrastructure

will fi ll out the alleged BB gap that at present exists between Europe and

North America. But lack of entry and investment in the local exchange

has raised doubts about the effi cacy of this policy. Critics argue that the

regulatory approach which initially favors competition on service via an

unbundling mechanism does not necessarily develop to a facilities compe-

tition at later stages (the ladder of investment theory). Regulators should

consider new frameworks to promote investment and innovation in the

industry.

The goal of this chapter is to present the theoretical underpinnings

of the ladder of investment theory and to review the recent literature

on the eff ects of the ladder of investment on BB performance of major

industrialized countries. By doing so, we aim at shedding more light on

the issues concerning the role of regulation in the era of rapidly con-

verging technologies. Our analysis and the review of recent empirical

studies reveal that the ladder of investment regulation is not fl awed per

se, but the way it is applied in various countries may lead to diff erent

BB performance (Japan versus Australia, for instance). It seems that the

regulatory agencies have to develop more refi ned techniques and strate-

gies which give incentives to new entrants to become fi rst- movers in the

investment race.

Section 1.2 of this chapter presents the BB market and the technologi-

cal platforms that exist to provide BB services. The institutional settings

of the broadband market are examined and this sets the debate as far as

investment in BB technologies is concerned. Section 1.3 reviews the main

arguments by examining the relationship between regulation and invest-

ment in broadband infrastructure. It examines critically the arguments

according to which innovation and investment in broadband infrastruc-

ture may be promoted within a regulatory environment which favors

initially a service- based competition (SBC) and later a facilities- based

competition (FBC), the so- called the ladder of investment theory. Section

1.4 reviews the fi ndings of the empirical studies, particularly those dealing

with the ladder of investment theory. The purpose of this section is to

critically examine the robustness of the results of the most recent empirical

studies and their usefulness to regulators and policy- makers in gaining a

better understanding of the functioning of the investment behavior under

alternative regulatory approaches. Section 1.5 concludes and off ers some

policy recommendations.


1.2 TECHNOLOGICAL PLATFORMS FOR INNOVATION AND INVESTMENT IN BROADBAND: INSTITUTIONAL SETTINGS AND TECHNOLOGICAL PERSPECTIVES

There are currently several technologies capable of providing access to

broadband services each one with diff erent technological characteris-

tics. They can be divided into wired access technologies such as ADSL,

ADSL2+, VDSL, hybrid fi ber coaxial (HFC) and fi ber; and wireless

access technologies such as Wi- Fi, WiMAX (Worldwide Interoperability

for Microwave Acess) and satellite. From a customer’s point of view, the

most important characteristic of these technologies is speed. A number

of emerging technologies provide high speed broadband of 10Mbps and

higher, such as optical fi ber, HFC using DOCSIS- 2 (Data Over Cable

Service Interface Specifi cation) standard, VDSL (very high speed digital

subscriber line) and fast wireless broadband. These alternative broadband

technologies have been promoted as would- be competitors in diff erent

markets. However, competition across platforms is increasingly simmer-

ing through the evolving technologies. New entrants and incumbents may

use one or more platforms to off er their services but high speed broadband

requires upgrade of the fi nal link from the exchange to end- users (the

so- called ‘last mile’). These upgrades increase the quality of the network

and allow the off er of broadband services at speeds well in excess of those

off ered by existing technologies. The upgrades may be done on the existing

copper wires but both incumbents and new entrants may invest important

amounts in rolling out FTTH (fi ber to the home) networks. The new net-

works free the ‘last mile’, that is, the copper part which is still considered

as a bottleneck, and off er more bandwidth.

Competition from cable operators is becoming increasingly intense par-

ticularly by employing strategies aiming at extending the potential of their

HFC networks. The upgrade of their networks to higher bandwidths with

the deployment of DOCSIS- 3 standards necessitates migration to new

topologies and networks. The competition possibilities are quite high but

the investments to be realized in line with current network infrastructure

and new networks are quite substantial and risky and this makes both

entrants and incumbents reluctant to invest. Since current investments

in infrastructure are considered inadequate to provide ubiquitous broad-

band, actual competition is limited and potential competition is latent. In

such a context, interplatform competition may have little impact on local

markets and the existing institutional frameworks. Regulatory authorities

should then work on the elaboration of more appropriate conditions that

will make latent competition eventually eff ective. The current regulatory


framework adopted in many countries which favors competition by allow-

ing new entrants to lease the incumbent’s bottleneck capacity and off er

services to customers has been criticized as rather ineffi cient (Wallsten,

2006). The so- called local loop unbundling (LLU) regulation is based

on the assumption that as competition at the retail level becomes more

intense, competition at the wholesale level will eventually increase, pro-

vided that incumbents and entrants invest in infrastructure.

With LLU, competition at the retail level has indeed increased in some

countries (for instance, the UK; Ofcom, 2006) and became intense in

market segments where most services and applications are being devel-

oped. To lure customers, both incumbent and new entrants undertook

bundling strategies at the retail level (double, triple and quadruple play).4

To be sure, the regulatory framework of each country plays a signifi cant

role in the development of competition among players of converging

industries. For instance, due to technological and regulatory imperatives

cable operators off ered triple play in Europe and the US markets prior

to the telecom incumbents and Internet providers (Lee, 2009). This was

mainly due to the adoption of the Telecommunications Act of 1996 in

the US which gave the possibility to cable operators to establish telecom-

munications business. Further, Internet providers use bundling as a way

to access markets and to gain market shares from their competitors as

illustrated by the examples of Fastweb and Free (an Italian and a French

Internet provider, respectively) which managed to increase their market

shares with triple- play high- quality services.5

Moreover, telecom incumbents and mobile operators use bundling in a

defensive way, in order to protect their core market. The undertakings of

bundling strategies push all players to compete fi ercely in several markets

which further intensify competitive pressures. Diff erentiation of service

plays an important role in making each company unique in its off erings

and this allows it to preserve its market power. Based on evidence from

recent studies (Lee, 2009), companies manage to diff erentiate their pack-

ages by off ering at least one service in the package which is diff erent from

the same service in a competitor’s package.6

Despite the diff erentiation strategies, it is generally believed that the

intensifi cation of competition at the retail level would lead to more innova-

tion and investment in BB infrastructure7 (ladder of investment). But such

an outcome is not automatic. The experience with this approach through-

out the world is quite mixed. In many countries (Canada, Australia,

New Zealand, for example), competition at the retail level8 did not spur

adequate competition at the wholesale level and the adoption of the ladder

of investment approach left them behind others in terms of BB infra-

structure (see OECD, 2009; CRTC, 2009). Some specialists (Waverman


et al., 2007) have started questioning the (theoretical) foundations of this

approach and its practical applications. Others (Aron and Crandall, 2008)

have mostly questioned the ability of the regulatory agencies and the type

of regulation they have put in place to spur innovation. This research

purports to examine thoroughly the workings of the ladder of investment

theory and identify ways which could be used by the regulatory agencies

to spur BB investment. The next sections describe the ladder of investment

relationship in detail and explore the alternatives available to regulators to

prompt more investment in BB infrastructure.

1.3 REGULATION AND INVESTMENT IN BROADBAND INFRASTRUCTURE

The debate whether regulation is conducive to innovation is not new, but

recently has revived due to the apparent failure of regulation to provide

incentives to incumbents and new entrants to invest in infrastructure, par-

ticularly in broadband technologies. After many years of experience with

deregulation and the introduction of favorable regulatory conditions for

new entrants, which made competition a reality, economists and policy-

makers do not agree as to the level and kind of competition that is required

to induce telecommunication fi rms to invest in BB infrastructure. Many

doubt and some even blame (Aron and Crandall, 2008) the type of regu-

lation, that is, LLU, as the culprit of such performance. LLU and other

ambivalent regulatory policies have created an uncertain environment for

the incumbents and increased their systematic risk and the cost of capital,

as measured by their betas. This makes the conditions for accessing inter-

national fi nancial markets more diffi cult and jeopardizes their capacity

to get funding to invest in long- term projects.9 Recent statistics10 from

the Organisation for Economic Co- operation and Development (OECD,

2009) indicate the performance of major industrialized countries in BB

technologies (Figure 1.1).

Economists emphasize more and more the importance of dynamic

competition as a criterion of performance. For a long time, though,

static competition has been the criterion for performance and policies

were adopted aiming at promoting static effi ciency. The latter is achieved

when fi rms produce at their minimum long- run average production costs

(production effi ciencies) and consumers pay the marginal cost of goods

and services produced (allocative effi ciencies). In the telecommunica-

tions industry, regulators have designed various frameworks to achieve

static effi ciencies. Lately, these frameworks have been largely modifi ed

to accommodate dynamic effi ciencies. It is believed that static effi ciencies


could be achieved by promoting service- based competition. But it is also

hoped that service- based competition would eventually lead to facilities-

based competition (dynamic effi ciency). By allowing new entrants to

have equal access to the incumbent’s facilities, prices would eventually

be driven close to the cost of production and the social welfare would

be maximized, achieving thereby the static effi ciency objectives. Further,

lower prices for the telecommunications services would create suffi cient

new demand which eventually will be enough to provide incentives to

both incumbents and new entrants to invest in infrastructure and satisfy

this increasing demand. Thus static effi ciency would bring dynamic effi -

ciency which relates mostly to demand creation (new telecommunication

products and services) and innovation (new production techniques and

alternative technology platforms).11

In practice, the relationship between dynamic effi ciency and investment

seems to be non- linear. The combination of the so- called Schumpeterian

and escape eff ects may result in a reduction of innovation after a certain

threshold of competition. In reality, the two eff ects work in opposite

directions and the fi nal outcome depends on the size of each eff ect. More

precisely, as competition increases, it reduces the incentives of fi rms to

invest in new technologies or in new products because of the high risk of

post- entry reduction in rents (Schumpeterian eff ect). Nonetheless, at an

initial low level of competition it is possible to increase investment by pro-

moting more competition. As the latter increases so does the investment

by incumbents and new entrants. This is so because fi rms by investing

(innovating) manage to escape the negative eff ects of competition (escape

eff ect). Innovation is thus seen as the means to preserve (or even increase)

40

30

35

25

20

15

10

5

0

OECD

Other Fibre/LAN Cable DSL

Den

mark

Net

her

lan

ds

No

rway

Sw

itze

rlan

d

Icel

an

d

Ko

rea

Sw

eden

Fin

lan

d

Lu

xem

bo

urg

Can

ad

a

Un

ited

Kin

gd

om

Bel

giu

m

Fra

nce

Ger

man

y

Un

ited

Sta

tes

Au

stra

lia

Jap

an

New

Zea

lan

d

Au

stri

a

Sp

ain

Irel

an

d

Italy

Cze

ch R

epu

bli

c

Hu

ngary

Po

rtu

gal

Gre

ece

Slo

vak

Rep

ub

lic

Po

lan

d

Tu

rkey

Mex

ico

Figure 1.1 OECD broadband subscribers per 100 inhabitants, June 2009


economic rents. But as competition increases and surpasses a critical

threshold, a further increase in competition reduces investment and inno-

vation, the so- called Schumpeterian eff ect (Breschi et al., 2000; Malerba

and Orsenigo, 1996).

Accordingly, the initial level of competition, and its degree after entry,

determine the level of investment and innovation in the telecoms industry.

These combinations in conjunction with the variety of other instruments

used by the regulatory authorities to provide incentives to innovation

result in an array of outcomes which become diffi cult to disentangle. Thus,

although it is not easy to isolate them in practice, it can be said, in theory

at least, that countries may experience either an increase or a decrease in

investment and innovation in broadband technologies, depending on the

size of the Schumpeterian and escape eff ects. This result may be useful

to policy- makers and provide an important guidance to regulators in

the formulation of their policies. But to make it more operational and

to convert it to a powerful decision- making tool, optimality conditions

should be developed. It is therefore important to develop realistic models

which determine the threshold level of competition and provide answers to

the question of whether more competition – be in the form of free or low

access price, local loop unbundling, and so on – contributes to investment

and innovation in broadband technologies.

Empirical observation shows though that neither entrants nor incum-

bents have incentives to invest in infrastructure either in established or in

new technologies, and this despite the regulatory framework that favors

the service- based competition. It is not surprising, therefore, to see that

service- based competition is increasingly viewed as a source of dynamic

ineffi ciency and that regulators seek for new ways (total deregulation is

one) to bring incentives which would create an industry structure more

conducive to dynamic effi ciency.12 In New Zealand, a recent report (Boyle

et al., 2008) casts doubts that the telecommunications incumbents are

interested in investing in infrastructure, since if they do, their profi tability

will be reduced. It is stated that: ‘leaving the (investment) task to telcos,

they will ration technology and bandwidth to stretch out the life of yes-

terday’s investments and maximize their own returns’. According to the

same report, governments should, therefore, subsidize the investment in

broadband technologies, if an adequate level of broadband deployment

is to be achieved more quickly. In the same vein, some authors (Janssen

and Mendys- Kamphorts, 2008) go even further by arguing that: ‘there

may be a role for public investment and ownership of infrastructure, given

the seriousness of potential market failures’. Given these conclusions, it

becomes important to examine critically the underlying theory used by

regulators to formulate their recent regulatory policies and to investigate


thoroughly the conditions that lead to dynamic effi ciency. This is the

subject matter of the next subsection.

1.3.1 The Ladder Investment Theory of Regulation and Dynamic

Effi ciency

The dynamic impact of regulation on investment has been neatly articu-

lated as what is now called the theory of ladder investment (Cave, 2006).

According to the latter, the dynamic effi ciencies of competition could be

captured through a more or less slow process of deregulation and accom-

modation of entry. It is argued that entrants cannot invest in infrastructure

from the outset unless they get a customer base and familiarize themselves

with the costs, demand and other characteristics of the industry. To make

the entry attractive, regulators ought to reduce the risks of entry and allow

new entrants to realize some economic rents. At this stage, regulation

should aim at fi xing low access prices so that new entrants could access the

incumbent’s bottleneck facility. At initial low access prices new entrants

fi nd an interest in entering the market and competing head to head with

the incumbent on price while both entrants and incumbent use the same

infrastructure. At this point, entrants are not interested in investing in new

infrastructure but as they get familiar with the market and they realize

economic profi ts, their growth opportunities depend on their capacity

to retain proprietary essential assets. These proprietary assets represent

a competitive advantage and entrants have an incentive to invest in

infrastructure. Initially, service- based investments are complementary to

the technologies of the incumbent, but eventually new entrants and the

incumbent move up to the ladder of new investments by constructing new

facilities.

To attain this level of (facilities- based) competition, regulation plays

a signifi cant role. According to the ladder of investment theory, in order

to increase investment in infrastructure, regulators, after a certain period

of time, have to increase the access price, since its initial low level would

simply perpetuate the service- based competition. The higher access fees

would lower the revenues of new entrants unless the latter move on

the ladder and invest more in infrastructure. As the entrants gain more

know- how and master the technologies better, they can climb the ladder

even higher by investing more in new facilities. New entrants become

full- fl edged investors and owners of their infrastructure (Prieger and Heil,

2008), ready to compete with the incumbent not only in terms of price but

also in terms of technological and other demand-related characteristics.

Competition which started at the beginning as a service- based com-

petition (SBC) evolves to a facilities- based competition (FBC) where


incumbent and entrants vie for more market shares via prices and invest-

ments in new technologies (innovation). For regulators, the key to achieve

this outcome is to unbundle the local loop and allow new entrants to

access the most expensive part of the incumbent’s network. The local

loop unbundling (LLU) can take three diff erent forms: bitstream access,13

shared access14 and full unbundling.15 The terms and conditions for getting

one or another form of access depend on technological, proprietary, insti-

tutional and regulatory factors and the preferences of new entrants. For

instance, an entrant whose interconnection points with the incumbent’s

network are rather poor will prefer a bitstream connection, while entrants

with a more extensive network will prefer full access. Regulators should

then determine the terms and conditions of access by taking into account

the technological characteristics of new entrants and the type of competi-

tion that they desire to achieve in the telecommunications industry.

In sum, if regulators believe in the ladder of investment theory, they may

fi x very low access prices at the beginning. This discriminatory regulatory

regime provides incentives for further investment in infrastructure. To

achieve this goal though, the regulatory authorities have to develop a lot

of expertise in order to be able to determine what type of access to allow

and the timing of increasing the access fees. Further, regulators have to

take into account the strategic interactions of entrants and the incumbent,

but this is diffi cult for them particularly when they have to fi ne- tune and

coordinate competing interactions in a way to achieve the desired results.

Moreover, despite the regulators’ beliefs that the rise in access tariff will

incite new entrants to innovate and develop their own network, it is possi-

ble that the risks involved with new investments remain quite high, thereby

discouraging fi rms from delving into such investments. New entrants may

prefer to operate as simple resellers of traditional services. This point is

quite important because investment decisions are quite sensitive to risks

and the latter may be too high to make the investments profi table.

The main risks may be classifi ed as technological and fi nancial, but

other risks related to the market and to changes in the regulatory regimes

are also important. In a changing environment these risks are diffi cult

to quantify and as competition increases, the probabilities that the new

investments will provide economic rents are getting lower. This may be

true even though there would be an increase in the size of the market

resulting mainly from the introduction of new products and techniques

of production. Because competition reduces profi t margins, which are

important indicators of returns, fi nanciers are willing to provide fi nance

for projects having low risks and generous returns. Lack of fi nance is

an indicator that the investment projects are not highly valued by inves-

tors. Because competition is so keen and the risks are high, the telecoms


industry has been characterized as a mature, slow- evolving industry. Since

the initial development of the theory of the ladder of investment (Cave

and Vogelsang, 2003; Cave, 2006, 2007; Bourreau and Dogan, 2006) many

researchers have investigated in some detail the conditions under which

new entrants will invest in innovation and new facilities. The following

subsection presents the main arguments of the models.

1.3.2 The Optimal Timing of Innovation under the Ladder of Investment

Framework

The optimal timing of innovation under diff erent market structures and

regulatory regimes has been elaborated by Dobbs (2004) but his model

does not take into account the complexities of investment decisions when

regulators allow local unbundling under an asymmetric16 regulatory

framework. This is a serious drawback of the model because the behavior

of the incumbent and new entrants is signifi cantly altered when access is

regulated and the eff ects of this discriminatory regulation on investment

and innovation are unclear. A priori, it is expected that investment will

increase if new entrants can pre- empt the incumbent’s investment, but it

may decrease when rival fi rms cannot or are not allowed to pre- empt the

incumbent’s investment. This relationship is even less clear if we take into

account not only investment in new technologies but also investment in

existing technologies. Indeed, if investment decisions are lumped together

and no distinction is made between investment in fi xed line BB (existing

technology) and mobile BB (new technology), the results may be quite dif-

ferent. Given the increasing importance of mobile and other technologies

in delivering BB services, the investment decisions of both the incumbent

and new entrants may again become biased. It is thus appropriate to see in

more detail some of the models which tackle these questions. Their results

will be useful to regulators and other policy- makers given that current

regulatory policies in both Europe and North America weight the dynamic

aspects of competition more than the static ones.

Competition can take various forms but two of them are particularly

important for the ladder of investment theory, the SBC and the FBC.

Under the ladder of investment theory, the intertemporal variation of

access fees by the regulatory agency incentivizes new entrants to invest

in infrastructure. When competition is in the form of SBC, the entrants’

initial investments are relatively small, since they use the incumbent’s

network. Nonetheless, they have to make an irreversible investment and

sink an initial amount of funds, ISBC, into complementary technology.

But as the intensity of competition increases and the regulatory agency

increases the interconnection fees, entrants have to make IFBC irreversible


investments in new facilities which are much higher than the initial invest-

ments in SBC. If the time to make the irreversible investment when the

fi rm is under SBC is tS, and the time to make the irreversible investment

when the fi rm is under FBC is tF, where tF > tS, and if CFn are the net cash

fl ows of the investment of new entrants, then the expected market value of

new entrants will equal the sum of discounted value of the investment and

the net cash fl ows that the investments will bring over time.17

The regulated access fee plays a signifi cant role on the timing of

investments. On the one hand, investment in SBC is determined by

the regulator since it is up to him to specify the type of access that the

incumbent must off er to new entrants. On the other hand, the regula-

tor, by determining the level of access fees, determines new entrants’ net

cash fl ows and their capacity to attract funds to invest in SBC. Since

the decision to invest can be viewed as a real option18 for new rival

fi rms, the latter provides a tool to fi rms to manage the risk arising from

the investment, but this option also provides a bias to the investment

decision. Since rivals will invest only if they are sure that their invest-

ment will be profi table, it will be more interesting for them to wait and

see what happens to the investment realized fi rst by the incumbent. If

the incumbent’s investment is profi table, rivals will invest too, but if

the incumbent’s investment is not profi table they will prefer to wait.

Thus during good periods investment will increase but this reduces the

potential profi tability of the incumbent (and new entrants), and during

bad periods investments are retarded and all the risks are assumed by

the incumbent. Regulation provides an asymmetry in sharing the risks;

the incumbent bears the full risk in bad periods but in good times the

proceeds are shared by the incumbent and entrants alike. Given that the

incumbent’s cost of capital is the opportunity cost during good and bad

times, the incumbent would not be interested in investing either, or at

least its commitment to invest would be relatively low (see Table 1.1).

The situation becomes even more complex when we take into account

the level and kind of competition that will prevail on the retail market.

The level of competition on the retail market determines the revenues that

the incumbent will receive from the wholesale market. If the incumbent is

quite aggressive in the retail market this will have an impact on the rev-

enues of new entrants and the incumbent alike. If new entrants are losing

market shares, the reduction of their revenues has a direct impact on the

incumbent’s capacity to safeguard a steady cashfl ow from the business

line of new entrants.19 Low access price and high retail prices make invest-

ment in SBC more attractive than in FBC. Entrants delay investments in

infrastructure, and competition in facilities does not develop as regulators

have originally expected.


Determining the timing of investment is quite important for regulators

and policy- makers alike. Further, the timing depends on the game the

incumbent and new entrants play. But the results are sensitive depend-

ing on which fi rm invests fi rst, the incumbent or the new entrants.

Stackelberg- type models have been developed to illustrate the intricacies

of competition at the retail level and its results on investment. There are

undeniable advantages for each fi rm to be a leader in the industry in

terms of investing fi rst in BB technologies. Obviously, the most important

fi rst- mover advantage is monopoly profi t. The leader reaps monopoly

profi ts as long as a follower does not replicate the leader’s investment.

But there are benefi ts associated with the second- mover too. The second-

mover advantages may be a better understanding of new technologies,

an eventual reduction of the costs of followers, a discovery of new prod-

ucts supplied by the new technologies, and so on. Thus, being second in

the race for investment in BB infrastructure can bring some undeniable

benefi ts.

The equilibrium in the market is not unique but depends on the new

entrants’ preferences concerning their position in the race for invest-

ment (being fi rst or second). There are two diff erent types of equilibrium

depending whether new entrants prefer the benefi ts of the second- mover

or are indiff erent between the advantages of fi rst- mover or second- mover

(Keiichi Hori and Mizuno, 2009). In the absence of an obligation to

invest within certain time limits, the preferences of new entrants deter-

mine the outcome of the game. If new entrants are of a second- mover

Table 1.1 Pay- off s and the timing of investment under the ladder of

investment regime

Regulated access fees and

timing of adjusting them

ISBC

is less inductive to

investment

IFBC

is more inductive to

investment

Period 1: tS Period 2: tF

High (H) N/A Climb the

ladder of

investment

Incumbent

invests

[E(MV)]>0

(expected

market

value)

entrants

invest

Low (L) Incumbent

invests in

upgrading

Entrants

invest

marginally

N/A N/A

g

g


type, investments will be delayed and the introduction of competition

by the regulatory authorities does not have a real impact on the timing

decisions to invest by the incumbent.20 In the opposite case, when new

entrants are of fi rst- mover type, investments will be accelerated and the

introduction of competition by the regulatory authorities does have a

real impact on the incumbent’s timing decision to invest. These two equi-

libria, known as the ‘waiting game’ and the ‘preemption game’ (Guthrie,

2005), are quite useful for regulatory purposes.

Nonetheless, it is not easy for the regulatory agencies to identify the

type of new entrant (fi rst- mover or second- mover), nor the reactions of

the incumbent. Given that the models do not provide a clear- cut solution

to the problem faced by the regulatory agencies which want to promote

dynamic effi ciencies, it is important for them to be proactive and develop

strategies that make new entrants fi rst- movers in the investment decision

process.21 It is probably the only way to bring more investment in BB tech-

nologies and accelerate the growth rates of the economy in general and the

telecommunications industry in particular.22

1.4 INTERNATIONAL REGULATORY APPROACHES TO INVESTMENT IN BROADBAND INFRASTRUCTURE

Although it is impossible to present each country’s regulatory approach to

investment in broadband, it is advisable to review some of them in order to

get a comparative perspective. It is important to stress from the outset that

in some countries facilities- based competition is non- existent and/or quite

limited since there is only one platform to provide broadband services,

usually through DSL.23 In other countries, such services can be provided

on an interplatform basis via DSL and cable modems. In either case, an

upgrade of the networks is necessary in order to be able to provide broad-

band services. The old telephone and cable networks have been developed

to be shared by a certain number of users and a particular traffi c fl ow (bidi-

rectional in the case of telephone services but unidirectional for cable).

Cable companies have to upgrade their network to accept a bidirectional

traffi c fl ow and telephone companies have to upgrade their network to

increase their transmission performance, especially for the ‘last mile’ (fi ber

to the home) for services such as video which requires more bandwidth.

Also, to get closer to customers’ premises and shorten the length of the

local loop, the telephone companies need to increase the number of central

offi ces. This is also a way to strengthen the transmission performance of

their DSL modems.


The existence of alternative technologies and vintages has aff ected both

the regulatory and the institutional environments in which the communi-

cations incumbents have evolved. In the case of cable- DSL environment,

the regulation has been asymmetric, in the sense that cable companies

enjoyed less stringent regulations than telecommunication fi rms because

the former were more competitive due to the existence of alternative

platforms such as satellite and terrestrial transmission which were judged

suffi cient to keep monopoly power at bay. By contrast, the telecommu-

nications industry has always had a more rigid regulatory environment

because of the existence of one technology platform (copper and coaxial

cables) which was considered a source of ‘market power’. Incumbents

could use it to reap consumer surpluses. From the consumers’ interest per-

spective, regulatory approval of retail prices and control of entry were thus

judged necessary. This historical evolution and country particularities

are refl ected in the diversity of approaches adopted by various countries

to promote investment in BB infrastructure. But behind each approach

there is a common hypothesis: regulatory policies which initially focus on

fostering competition at the service level will eventually provide incentives

to both entrants and incumbents to invest in BB infrastructure. Service-

based competition moves to a higher level and it becomes facilities- based

(ladder theory).

Many countries have adopted diff erent regulatory frameworks. As a

consequence, their ranking in terms of BB penetration diff ers signifi cantly

(OECD, 2009). The diff erences in performance may be explained by many

factors,24 but the most recent studies have focused on the institutional

setting which prevails in each country, particularly access regulation and

LLU. To off er guidance to policy- makers and regulatory authorities, we

briefl y summarize the main fi ndings of these studies. Most of the studies

have been realized on an aggregate level. They examined the relationship

between access regulation and investment in BB infrastructure in order

to off er guidance concerning the use of the best international practices to

achieve the objective of dynamic effi ciencies.

For instance, an OECD (2007) study examined empirically the eff ects

of unbundling on the rollout of BB technologies. This was the fi rst sys-

tematic empirical attempt to test this relationship using data for 2002 and

2005 from a number of OECD countries. The conclusions were very pow-

erful and quite straightforward. It appeared that local loop unbundling

was a catalytic factor for the deployment of BB in the OECD countries

examined. The study concludes that: ‘unbundling . . . is currently more

signifi cant than platform competition in explaining broadband penetra-

tion’ (p. 20). The author of the OECD study underlines the important role

that the national regulatory authorities (NRAs) can play in determining


an appropriate structure of access fees for the unbundled local loop infra-

structure and in using subsidies for more BB penetration.

These results were in sharp contrast to those obtained in previous

empirical studies. For instance, a group of studies (Distaso et al., 2006;

Cava- Ferrer and Alabau- Munoz, 2006; Kim et al., 2003) found that

the eff ects of unbundling on BB rollout, although positive, were very

small but statistically insignifi cant. Another group of studies (Denni and

Gruber, 2005) found that the eff ects were rather transitory and very small,

while others (Wallsten, 2006) found that these eff ects were not always posi-

tive or statistically signifi cant. This discrepancy in the empirical results led

Boyle et al. (2008) to investigate further both the data used in the OECD

study and the econometric techniques.

Indeed, Boyle et al. (2008) vehemently criticized the OECD results on

both economic and statistical grounds. First, they argue that unbundling

is just one single factor but there are many others which may aff ect the

BB uptake. Singling out the LLU factor ignores other important factors

which may contribute to the explanation of the relationship between

access regulation and BB investment. Second, once they correct for the

estimation procedure used in the OECD study, the ‘statistically signifi cant

relationship between local loop unbundling and broadband uptake disap-

pears’ (p. 4). The authors conclude that there are neither economic reasons

nor statistical ones to justify the strong relationship between unbundling

and BB rollout. This refuting of the OECD results brings the debate back

to its fundamental question concerning the role of access regulation in the

deployment of BB technologies.

In another study, Gruber (2007) examined the relationship between

regulation and investment in the EU and he found that new entrants

have a tendency to use a single technological platform and also to avoid

investing in infrastructure. The author also investigated the behavior

of entrants and incumbents in mobile technology. He asserts that the

behavior of entrants and incumbents in this sector is in stark contrast

with that observed in fi xed lines infrastructure. In the mobile sector,

competition on alternative technological platforms is already quite

widespread and access regulation is less prevalent. Investments in BB

infrastructure in the mobile telecommunications sector are more aligned

with the market shares of new entrants and incumbents. In order to

favor a greater expansion of the BB technologies, he proposes a new

regulatory framework which puts greater emphasis on platform competi-

tion through incentives and special access regulation provisions. His ad

hoc analysis shows that the investment behavior in the mobile sector

is indeed aligned to the ladder of investment theory, but the fact that

the behavior of incumbents and entrants in the fi xed line sector is quite


diff erent than expected reveals that the origin of the investment problem

resides in access regulation. He cites the FCC’s (Federal Communication

Commission) successful shift of focus from the regulation of broadband

access and wholesale markets towards more infrastructure competition.25

All in all, he concludes by stating that: ‘the underlying hypothesis of the

current regulatory regime, i.e., appropriate access prices lead to facility

based competition, still awaits confi rmation’. He emphasizes the need to

adopt ‘new measures which could be much more conducive to new infra-

structure investment’ (Gruber, 2007, p. 23).

In the fi rst exhaustive empirical study concerning the link between

dynamic effi ciencies, regulation and investment in BB in Europe, Distaso

et al. (2009), examined the approaches adopted by 12 European regula-

tory authorities to increase investment in BB infrastructure. They claim

that the approaches adopted by national regulatory authorities (NRAs)

were mostly consistent with the ladder of investment theory, but given

the complexities of national markets, the results could not be interpreted

properly. By dividing the sample into three subgroups,26 they were able to

make comparisons among individual countries within each category and

identify the ones which have adopted the highest level of competition and

diff usion of BB services. They found that although considerable progress

has been realized, the adoption of more dynamic pricing schemes at the

wholesale level and/or commitments by the regulatory agencies stipulating

the phase- out period of asymmetric regulation is still so slow that it makes

it diffi cult to predict ‘when and in which way regulators will introduce the

full set of ladder of investment policies’ (Distaso et al., 2009, p. 13). This

study is the fi rst one to examine empirically the above- stated relationship

but it falls short of giving more guidance to regulators in terms of the elab-

oration and adoption of a better regulatory platform to increase invest-

ment in BB infrastructure. It is true that the limited sample and incomplete

data delimit the scope of the analysis. Obviously, additional detailed data

are needed to get more convincing results.

Sutherland (2007) analyzes the eff ects of unbundling on BB deployment

in several countries and he concludes that countries which have followed

the successful pattern of infrastructure competition perform better than

those which have chosen the service- based competition. Like the previous

authors, he does not distinguish between investments in fi xed or mobile

networks and therefore he cannot make a precise statement about the

eff ectiveness of the regulatory regime which puts the unbundling at the

center of BB deployment.

All in all, the theoretical underpinnings of the ladder of the investment

theory are quite solid but the empirical evidence does not show clearly that

it can work in practice as the theory predicts. Despite the limitations of the


empirical studies, it can be said that there are some indications that the

existence of alternative platforms and the competition that they entail are

more conducive to investment in BB infrastructure than the service- based

competition. Nonetheless, in both cases, regulation plays a signifi cant role

because it defi nes the appropriate conditions for both types of competi-

tion to develop. The regulators’ role is thus enhanced during this phase of

deregulation and reliance on market forces.

1.5 CONCLUSIONS AND POLICY RECOMMENDATIONS

The level of broadband investment is not the same across countries. Given

the importance of BB to economic growth and development, various

regulators have adopted original regulatory frameworks to promote

investment in BB infrastructure and safeguard the benefi ts of dynamic

competition. The ladder of investment theory has contributed to a great

extent to the development of these new regulatory frameworks. According

to this theory, regulators by favoring LLU and other types of asymmetri-

cal regulation (which favors new entrants rather than incumbents), are

able to reap initially the static benefi ts of competition and eventually the

dynamic ones.

To be achieved, the regulatory agency has to gradually adjust the access

fee entrants pay to get connected to the local loop. The increase of access

fees makes service- based competition less interesting in terms of profi ts

and this gives incentives to entrants to make investments in new facilities.

According to this theory, it is possible to achieve the desired dynamic effi -

ciencies by using a regulatory framework which appropriately regulates

the access fees. Our theoretical analysis reveals that the ladder of invest-

ment regulation is not fl awed per se, but the way it is applied in various

countries may lead to diff erent BB performance as the cases of Japan and

Australia make clear. The empirical studies reviewed in this chapter dem-

onstrate that the application of the ladder of investment theory in practice

did not bring the anticipated benefi ts of dynamic effi ciency. New entrants

have either been passive in terms of investment in new infrastructure or

their investments fell short of expectations. Incumbents too have limited

their investments in BB technologies by making small upgrades just to

keep up with the evolving demand for bandwidth. It seems that the regula-

tory agencies have to develop more refi ned techniques and strategies that

give incentives to new entrants to become fi rst- movers in the investment

race. By this process, the competition for infrastructure will bring the

benefi ts associated with dynamic effi ciencies. Thus, the role of regulation


may be even more important now that the telecommunications industry

has became more competitive than it used to be, when competition was

rather limited.

NOTES

1. We would like to thank the anonymous referees for their valuable comments and sug-gestions. We also thank Stephen Schmidt and Anders Henten for reading the chapter and making very insightful comments. All errors are ours.

2. Investment is not necessarily synonymous with innovation. Since our emphasis here is on broadband technologies which are new innovative technologies, we use both terms interchangeably.

3. Wireline access is not necessary to off er mobile broadband, which may be done on a full- fl edged facilities- based competition.

4. At the retail level, bundling refers to the practice of selling two or more diff erentiated products as a package.

5. Free’s monthly fee, €29.99, has never increased since 2002, and its ‘triple- play’ services include very high- speed broadband Internet access (with downloads up to 24 megabits per second and uploads up to 1 Mbps), unlimited calls (to France and 70 countries abroad) and more than 100 digital television channels. Its decision to adopt point- to- point FTTH will allow Free to deploy in a cost- eff ective way fi ber optical links to each home and enable it to off er 1 Gbps and even 10 Gbps connections over the same fi ber (Cisco, 2008).

6. In the case of cable and telephone companies, their diff erences in the triple- play bundle were mostly on the video service category while voice and data was rather similar (Lee, 2009).

7. Ofcom has stated that: ‘Sustainable competition is only one element necessary for an eff ective broadband market . . . However, we recognize that competition is not the only issue for broadband customers and are working to address other important issues relat-ing to broadband separately’ (Ofcom, 2006).

8. In some countries, like New Zealand, it was the uncertainties concerning whether mandatory LLU would have been adopted by the regulatory authorities that left this country behind its major trading partners in terms of BB infrastructure.

9. Aron and Crandall (2008) have estimated the betas (systematic risk) of major incum-bent local exchange carrier (ILECs) and found that the most recent two- year betas (2006–08) are higher than the fi ve- year betas (2003–08) except for France Telecom.

10. These statistics have been the subject of heated debate lately because they fail appar-ently to measure accurately the BB infrastructure and by doing so they result in a biased ranking of the countries concerned.

11. The static–dynamic effi ciency is at the forefront of the new European regulatory frame-work for electronic communications services. It is stated explicitly there that regulation should ‘promote competition’ and ‘encourage effi cient investment in infrastructures and to promote innovation’ (European Parliament and Council of the European Union (2002), p. 38.

12. Some economists (Wallsten, 2006) do not see it as a failure of competition but, on the contrary, as a failure of asymmetrical regulation which favors new entrants at the expense of incumbents.

13. Bitstream access is the most basic form of access to the incumbent’s network. Under this type of access, entrants cannot modify any of the technical characteristics of the technology used by the incumbent by adding devices or other equipment, and they are de facto restricted to the function of reselling. Since new entrants have a very limited infrastructure themselves and quite a few interconnection points with the incumbent’s


network, it is not surprising to see that the best strategy for new entrants is to rent at a bulk price the incumbent’s services and resell them to fi nal customers.

14. Shared access allows entrants to off er services to their customers other than voice by leasing the high frequency non- voice spectrum of copper wire from the incumbent. The latter owns and controls the entire line and off ers voice services normally by using the low- frequency voice spectrum.

15. With fully unbundled access, entrants get hooked to the incumbent’s capacity by gaining full control of the copper line and this allows them to off er a full gamut of serv-ices to their customers competing head to head with the incumbent.

16. The asymmetries arise because the regulator favors new entrants by fi xing low- access fees while regulating the incumbent on a number of monopoly and potentially competi-tive segments of the market.

17. E[MV(tF, tS)] = E c2 ISBC

(1 1 r) tS1 a

tF

t5tS11

CFn

(1 1 r) t2

IFBC

(1 1 r) tF1 a

`

t5tF11

CFn

(1 1 r) td .

This equation does not take into account any interactions among incumbents and new entrants and ignores their strategic behavior.

18. Real options have been studied extensively in the literature but this approach has still not managed to give more precise answers to the questions raised here (Smit and Trigeorgis, 2004).

19. Of course, this problem may be avoided or at least become less acute in a growing industry.

20. Obviously, the incumbent’s cashfl ows are aff ected by the new entrants but its decision to invest sooner or later is not aff ected by entry.

21. Waverman and Dasgupta (2007), believe that: ‘the incumbent’s local loop infrastruc-ture will remain an enduring bottleneck . . . even if the current copper infrastructure is replaced by the Next Generation access network (NGNs)’. This enduring bottleneck warrants an ‘enduring regulation of the access network’ (ibid. p. 7).

22. Functional separation has been proposed as an alternative means to foster competition and investment in BB technologies. There are considerable doubts concerning the eff ec-tiveness of functional separation to bring the desired dynamic effi ciencies in the industry (Waverman and Dasgupta, 2007). Public–private partnerships (PPPs) have also been suggested as a means to increase investment in BB infrastructure.

23. This is the case for many European countries. For instance, Greece relies exclusively on DSLs to provide Internet and other communications services.

24. There is ample literature identifying various factors in explaining country performance in BB penetration. For a review of the literature see Gentzoglanis and Aravantinos (2008).

25. In 2003, the FCC (2003, 2005) shifted focus and phased out the obligation for unbun-dling because one of the incumbent regional telecommunications carriers won a legal case for refusal to unbundle.

26. The fi rst group of countries included those with high rates of bitstream and resale services and low investments in LLU and proprietary infrastructure. The second group included countries with high levels of LLU and shared access but low levels of propri-etary infrastructures. The third group included countries with high levels of proprietary infrastructures.

REFERENCES

Aron, D.J. and R.W. Crandall (2008). Investment in next generation networks and wholesale telecommunications regulation. White Paper, 15 September.

Bourreau, M. and P. Dogan (2006). ‘Build or Buy’ Strategy in the local loop. American Economic Review, 96(2), 72–6, May.


Boyle, G., B. Howell and W. Zhang (2008). Catching up in broadband regres-sions: does local loop unbundling really lead to material increases in OECD Broadband Uptake? NZ Institute for the Study of Competition and Regulation, July.

Breschi, S., F. Malerba and L. Orsenigo (2000). Technological regimes and Schumpeterian patterns of innovation. Economic Journal, 110, 388–410.

Cava- Ferreruela, I. and A. Alabau- Munoz (2006). Broadband policy assessment: a cross national empirical analysis. Telecommunications Policy, 30, 445–63.

Cave, M. (2006). Encouraging infrastructure competition via the ladder of invest-ment. Telecommunications Policy, 30(3–4), 223–37.

Cave, M. (2007). The regulation of access in telecommunications: a European perspective. Working Paper, Warwick Business School, University of Warwick, UK.

Cave, M. and I. Vogelsang (2003). How access pricing and entry interact. Telecommunications Policy, 27(10–11), 717–27.

Cisco (2008). French ‘triple- play’ service provider deploys fi ber to the home. Customer case study. http://www.cisco.com/en/US/solutions/collateral/ns341/ns524/ns562/ns577/case_study_C36- 454892_ns577_Networking_Solutions_Case_Study.html.

CRTC Communications (2009). Monitoring Report, Ottawa, http://www.crtc.gc.ca.

Denni, M. and H. Gruber (2005). The diff usion of broadband telecommunications: the role of competition. Paper presented at the International Communications Society conference, Ponte Verda, Spain.

Distaso, W., P. Lupi and F.M. Manenti (2006). Platform competition and broadband uptake: theory and empirical evidence from the European Union. Information Economics and Policy, 18(1), 87–106.

Distaso, W., P. Lupi and F.M. Manenti (2009). Static and dynamic effi ciency in the European telecommunications market: the role of regulation on the incentives to invest and the ladder of investment. University of Padova.

Dobbs, I.M. (2004). Intertemporal price cap regulation under uncertainty. Economic Journal, 114, 421–40.

European Parliament and Council of the European Union (2002). Directive 2002/21/EC of March 7 2002 on a common regulatory framework for electronic communications networks and services (Framework Directive) (Offi cial Journal (OJ) L108, 24.04.2002, pp. 33–50). Brussels.

FCC (2003). In the Matter of Review of the Section 251 Unbundling Obligations of Incumbent Local Exchange Carriers; Implementation of the Local Competition Provisions of the Telecommunications Act of 1996; Deployment of Wireline Services Off ering Advanced Telecommunications Capability, CC Dockets Nos. 01- 338, 96- 98, 98- 14, Report and Order and Order on Remand and Further Notice of Proposed Rulemaking, Federal Communications Commission, (Triennial Review Order).

FCC (2005). In the Matter of Unbundled Access to Network Elements; Review of the Section 251 Unbundling Obligations of Incumbent Local Exchange Carriers, WC Docket No. 04- 313 and CC Docket No. 01- 338 (Triennial Review Remand Order).

Gentzoglanis, A. and E. Aravantinos (2008). Forecast models of broadband dif-fusion and other information technologies. Communications and Strategies, Special issue (Nov), 73–98.


Gruber, H. (2007). European sector regulation and investment incentives for broadband communication networks. EIB Working Paper.

Guthrie, G. (2005). Regulating infrastructure: the impact on risk and investment. Journal of Economic Literature, 44(4), 925–72.

Janssen, M. and E. Mendys- Kamphorts (2008). Triple play: how do we secure the benefi ts? Telecommunications Policy, 32, 699–700.

Keiichi Hori, K. and K. Mizuno (2009). Competition schemes and investment in network infrastructure under uncertainty. Journal of Regulatory Economics, 35(2), 179–200.

Kim, H.S. Hong, J. Hee Kim, G. II Yoo and W. Ha Kim (2003). Emerging Broadband Access and IP Multimedia Architecture of KT. Technical Report, Korea Telecom 2003. Available at http://users.ece.utexas.edu/~hkim4/index_fi les/commag2003.pdf.

Lee, S. (2009). The triple- play bundle strategy of cable and telephone companies in the current US telecommunications market. International Journal on Media Management, 11(2), 61–71.

Malerba, F. and L. Orsenigo (1996). Schumpeterian patterns of innovation are technology- specifi c. Research Policy, 25(3), 451–78.

OECD (Organisation for Economic Co- operation and Development) (2007). Catching- up in broadband – what will it take? Communication Infrastructures and Services Policy Paper DSTI/ICCP/CISP(2007)8/FINAL, OECD, Paris. http://www.oecd.org.

OECD (2009). Organisation for Economic Co- operation and Development, Directorate for Science, Technology and Industry, OECD Broadband Portal. Available at http://www.oecd.org/document/54/0,3343,en_2649_34225_38690102_1_1_1_1,00.html.

Ofcom (2006). Review of the wholesale broadband access markets 2006/07. 21 November. http://www.ofcom.org.uk/consult/condocs/wbamr/summary/.

Prieger, J.E. and D. Heil (2008). Is regulation a roadblock on the information highway? In I. Lee (ed.), Handbook of Research on Telecommunications Planning and Management for Business (pp. 15–32). Hershey, PA: IGI Global.

Smit, H.T.J. and L. Trigeorgis (2004). Strategic Investment: Real Options and Games,Princeton, NJ: Princeton University Press.

Sutherland, E. (2007). Unbundling local loops: global experiences. Working paper, Link Centre.

Wallsten, S. (2006). Broadband and unbundling regulations in OECD countries. Working Paper 06- 16, June, AEI–Brookings JointCenter. Available at http://ssrn.com/abstract=906865.

Waverman, L. and K. Dasgupta (2007). Mandated functional separation: act in haste, repent at leisure? Unpublished paper, November.

Waverman, L., M. Meschi and M. Fuss (2005). The impact of telecoms on eco-nomic growth in developing countries. Unpublished paper.

Waverman, L., M. Meschi, B. Reillier and K. Dasgupta (2007). Access regulation and infrastructure investment in the telecommunications sector: an empirical investigation. Unpublished paper.

43

2. Intermodal telecommunications competition: implications for regulation of wholesale services

William E. Taylor

2.1 INTRODUCTION

The historical paradigm for regulation of wireline telecommunica-

tions services has been to regulate both wholesale and retail services of

incumbents with the expectation that continued regulation of wholesale

services – mandatory unbundling and provision of essential wholesale

services at regulated prices – might someday encourage entry and

investment in facilities that would permit deregulation of retail services.

Regulation of wholesale services would remain a necessary feature of

the landscape, unless competition develops for wholesale services. As

it happens, intermodal competition for retail telecommunications serv-

ices has turned this expectation on its head. Paradoxically, intermodal

competition for retail services makes regulation of wholesale services

unnecessary.

Intermodal competitors for retail services such as cable, wireless and

Voice over Internet Protocol (VoIP) suppliers serve their customers

without requiring use of incumbent local exchange carrier (ILEC) facili-

ties, and retail telecommunications markets are often eff ectively com-

petitive. In this case, there is no economic rationale to regulate wholesale

services. The demand for wholesale services is a derived demand, derived

from the demand for retail telecommunications services. And if depend-

ent wireline competitive local exchange carriers (CLECs) can earn no

supracompetitive returns in the retail market, so an ILEC monopolist of

wholesale wireline services cannot increase its profi ts by charging a supra-

competitive price for its essential wholesale services.

Where intermodal competition, by itself, is insuffi cient to constrain

retail prices but requires competition from dependent wireline CLEC

competitors, an unregulated ILEC might be able to exercise market

power over wholesale services. In this case, the wholesale service might be


considered ‘essential’ in the sense that competitors require it to compete

in the downstream market and there is some likelihood of ILEC anticom-

petitive conduct in the retail market. Even in this case, however, ex ante

regulation in the wholesale market would likely be self- defeating. First,

without a market trial, we would never know what market- based whole-

sale prices might be and what competition would emerge at those rates and

become sustainable in the retail markets. Second, the mandatory unbun-

dling of ILEC facilities at regulated rates would reduce the incentive of

retail wireline competitors to invest in their own network infrastructures

and compete on an end- to- end basis. At the same time, the requirement

that ILEC facilities be shared with competitors reduces the ILEC’s incen-

tives to introduce and roll out that infrastructure, particularly for services

associated with investment that will be – but is not currently – sunk. Third,

price regulation at multiple stages of production – wholesale and retail –

is particularly problematic, inviting ineffi cient arbitrage and unintended

substitution.

The institutional and market setting of this chapter is the US. However,

the ubiquity of technology – particularly the growth of intermodal alterna-

tives to traditional wireline telephony – means that these conclusions are

applicable more generally.

2.2 BACKGROUND

Stemming from the assumption that parts of the telephone network are

natural monopolies or essential facilities, regulation of wholesale services

can make competition possible for retail services. Terms and conditions

and prices of wholesale services are pervasively regulated, as there are not

generally multiple suppliers of these facilities. To be clear, retail services

are the familiar telecommunications services that residence and business

customers buy: access to the network, usage, and ‘vertical’ services such

as call answering or automatic call forwarding. Wholesale services are

sold by one carrier to another, generally to permit the purchaser to supply

telecom services to retail customers. There are four kinds of players to

keep track of in these markets:

1. ILECs, which supply wholesale and retail services using traditional

wireline technologies. Examples include Verizon, AT&T and Qwest in

the US.

Competitive local exchange carriers (CLECs), which use traditional wire-

line technologies1 and come in two relevant fl avors:

Intermodal telecommunications competition 45

2. dependent CLECs, which use ILEC facilities to serve their customers;

and

3. facilities- based CLECs, which build or buy (from non- ILEC carriers)

their own facilities.

Obviously, many CLECs serve customers with their own facilities where

such investment is profi table and with ILEC facilities where it is not.

Finally,

4. Intermodal carriers that do not use ordinary wireline technologies to

reach their customers. Examples include cable companies (Comcast,

Cox), wireless carriers (Verizon Wireless, Sprint) and VoIP carriers

(Vonage).2

2.2.1 Historical Retail and Wholesale Regulation

Privatization of telecommunications companies since the mid- twentieth

century has generated a wealth of experience in regulating and deregu-

lating retail telecom markets in the US, the UK, the EU and elsewhere.

Pervasive cost- of- service regulation has evolved through price regulation

to deregulation or pricing fl exibility regimes as competitive circumstances

in the retail markets warranted. In this process, there has been general

agreement that absence of market power in the retail market is the relevant

trigger for reduced regulation.

For wholesale regulation, the history is shorter and the guiding princi-

ples ambiguous. US policy since the breakup of the Bell System in 1984

assumed a single vertically integrated ILEC network in each geographic

market with dependent competitors that required interconnection and

access to customers. This model has produced two radically diff erent

regulatory regimes. For long- distance carrier access services, regulation

by the Federal Communications Commission (FCC) (and state regulatory

authorities) has evolved along traditional patterns: cost- of- service regula-

tion replaced by price caps with some very limited pricing fl exibility for

interstate special access services. In contrast, for access to the local network

to provide local services, regulation under the Telecommunications Act of

1996 has followed a diff erent path. ILEC provision of unbundled network

elements for local service is required wherever CLECs would be impaired

without them, and wholesale prices are set at the long- run incremental cost

of a hypothetical effi cient supplier of network elements.

For both paradigms, with the exception of pricing fl exibility for special

access services, there has been little consideration of deregulating prices or

mandatory provision of these wholesale services.


The purpose of this wholesale regulation was not simply to control

market power but also to induce self- sustaining competitive entry into

the retail telecommunications markets.3 In theory, new entrants would

evolve from dependent to facilities- based competitors by investing in their

own facilities as their customer bases grew.4 If such an evolution took

place, then CLECs would no longer be ‘impaired’ without access to ILEC

unbundled elements at regulated prices.5 Thus, mandatory supply and

regulation of ILEC wholesale services would be expected to diminish as

the ability of CLECs to compete without using the ILEC’s network ele-

ments increased.

Similar models apply outside the US. In many countries,6 traditional

monopoly wireline ILECs are required to supply wholesale telecommuni-

cations services to their competitors, and terms and conditions (including

mandatory access and pricing) for such wholesale services are pervasively

regulated.7 Moreover, in determining whether regulation of wholesale

services is necessary, many regulators have defi ned separate wholesale and

retail markets and have examined market power separately in the whole-

sale market.8

Two surprises occurred. First, the ‘ladder of investment’ theory failed to

predict market outcomes. Few wireline CLECs evolved into full facilities-

based carriers, particularly for mass- market residential and small business

customers. CLECs grew quickly as long as the ILECs’ unbundled network

element platform (UNE- P) wholesale service was available at TELRIC-

based prices.9 However, once these services were no longer available,

CLEC growth stopped, and by June 2008, the majority of facilities- based

CLEC access lines belonged to cable companies – which never used ILEC

facilities – rather than to the traditional wireline CLECs. Second, retail

markets became more competitive due to intermodal carriers that made

no use of ILEC facilities. The ILEC and CLEC combined share of local

access connections fell steadily as demand shifted to intermodal suppliers

such as wireless, cable and VoIP suppliers.

2.2.2 Future Regulation

In light of these surprises, it is timely to ask whether continued regulation of

wholesale services is necessary and what standards we should use to assess

the wisdom of deregulation. At issue are the circumstances under which ex

ante regulation of unbundling requirements and pricing of wholesale serv-

ices is preferable to ex post regulation, in which the ILEC would be under

no explicit obligation to provide wholesale services or price them at some

regulated rate, leaving regulatory or competition authorities to respond ex

post to specifi c complaints of anticompetitive behavior.


Regulation entails costs and benefi ts. The potential economic benefi ts of

ex ante economic regulation of wholesale and retail services are fundamen-

tally diff erent in economic theory. As discussed below, economic welfare

eff ects are measured in the markets for fi nal goods – the retail markets –

and it is only the downstream eff ect of regulation at the (upstream) whole-

sale level that matters. As for costs, the cost consequences of regulating

wholesale markets are more complex.

First, regulation at successive stages of production (that is, of wholesale

and retail markets) can be inconsistent and a source of important unin-

tended consequences in competitive markets. A modest example from

the US: the FCC required that wholesale unbundled loops be priced at a

version of incremental cost, which was disaggregated into at least three

diff erent geographic areas to refl ect cost diff erences among urban and

rural exchanges. On the retail side, many states required that the price of

basic exchange service refl ect ‘value- of- service’, so that retail prices were

higher in urban exchanges (where there were more customers in the local

calling area) than in rural exchanges. The resulting regulation of both

wholesale and retail prices led to ineffi ciently large margins in urban areas

and ineffi ciently small (and negative) margins in rural areas.10

Second, regulation at the wholesale level inevitably induces distor-

tions in the retail markets because some network platforms are regulated

and others are not.11 Such regulatory disparity is not a simple squabble

over rents; it does not merely transfer wealth among competing carriers.

Rather, because telecom markets are characterized by rapid technologi-

cal change and competing platforms that are subject to lock- in or path-

dependence, regulatory disparities can have large and irreversible welfare

eff ects on consumers. Platforms currently in play include fi ber to the

home, fi ber to the node, coaxial cable, various incompatible mobile wire-

less platforms and various fi xed wireless alternatives. Regulatory distor-

tions here are inevitable, if only because the regulators’ jurisdiction diff ers

across the wireline, cable, wireless and broadband platforms. And distort-

ing a competitive market outcome here could drive the market to adopt an

ineffi cient platform or technology which could then persist for years.

2.3 INTERMODAL COMPETITION

Historically, diff erent networks were designed and deployed to carry dif-

ferent types of traffi c. The wireline public switched telephone network and

mobile telephone networks were optimized to transport basic voice com-

munications, cable networks were engineered to transport video, and the

Internet was designed to transport packet- based data traffi c. Today, these


technologies have converged so that providers can off er multiple types of

services over a single network. With convergence, the same services are

provided (and marketed) over diff erent network platforms (for example,

traditional cable systems, wireline telephone networks and wireless mobile

networks).

Three fundamental factors have driven this convergence: (1) technologi-

cal change (such as the advent of two- way, digital, broadband networks

and Internet Protocol – IP – technology) which has allowed all kinds of

wired and wireless networks to be used for any kind of service; (2) con-

sumer demand for bundled services; and (3) competition among providers

seeking gains from improved effi ciency (economies of scale and scope),

and the promise of increased revenues and lower churn rates. Because

convergence enables diff erent types of platforms to provide increasingly

similar bundles of services, traditional wireline carriers must now compete

with multiple platforms, including Internet and broadband services, cable

companies that have made substantial investments in their networks to

provide video, data and voice services, wireless services providers, VoIP

providers and other carriers using emerging technologies. These indus-

try developments have resulted in dramatic line losses to wireline local

exchange carriers and have made retail markets for telecommunications

services eff ectively competitive.

The growth of intermodal competition has been extensively docu-

mented in many forums.12 Here, I give some simple examples from US

data. Figure 2.1 shows dramatically the eff ect of intermodal competition

on local exchange access for US mass market telecommunications con-

nections. In June 2000, wireline access exceeded intermodal (wireless and

broadband) access by about 50 million lines; by June 2007, there were

approximately 200 million fewer wireline (ILEC plus CLEC) access lines

than intermodal connections.

Cable telephony is widely available across the US. By 2005, essentially

100 percent of US households had cable service13 and cable penetration

– by some measures – was 58 percent of homes passed.14 Comcast, for

example, reports that of its homes passed, broadband services were

deployed to 99 percent and telephone was enabled to 92 percent. Current

penetration of available homes was 30 percent and 13 percent for broad-

band and telephony respectively.15

Mobile wireless services are available throughout the US. Measured on

a census block basis, approximately 96 percent of the US population has

access to three or more wireless providers and more than half (57 percent)

have access to fi ve or more.16 Wireless subscribership approximately

doubled between December 2001 and December 2007, reaching about 250

million subscribers at the end of 2007. According to the FCC, about 187


million of these subscriptions were residential,17 so that residential wireless

subscriptions were approximately double the number of wireline sub-

scriptions in December 2007.18 That wireless service actually substitutes

for wireline service is implied by the growth of 8.4 million wireless- only

households since 2005; the latest National Health Institute Survey data

track the signifi cant growth in the percentage of adults living in wireless-

only households, a fi gure that reaches 17.5 percent of total households as

of June 2008 and has increased consistently and signifi cantly over time

across all age groups.19

Finally, for VoIP availability, high- speed Internet service is now avail-

able throughout the US. As of June 2007, about 96 percent of US Zip

Codes have three or more high- speed providers with lines in service, and

89 percent of all Zip Codes have four or more such providers.20 Digital

subscriber line (DSL) and cable broadband are both widespread: high-

speed DSL connections were available to 82 percent of the US households

where ILECs can provide local telephone service, while high- speed cable

modem service was available to 96 percent of the households where cable

system operators can provide cable TV service.21 Broadband subscription

rates have grown as well, reaching approximately 55 percent of US house-

holds in early 2008.22

A signifi cant proportion of US households use their broadband

0

50

100

150

200

250

300

350

Jun-00 Jun-01 Jun-02 Jun-03 Jun-04 Jun-05 Jun-06 Jun-07

Nu

mb

er o

f li

nes

or

sub

scri

ber

s, m

illi

on

s

CLECs

Wireless subscribers

ILECs

Wireless plus residential (small business) broadband

Note: Due to diff erences in reporting, 2005–06 data are not comparable to previous.

Source: US FCC (2007), US FCC (2009), Tables 1, 14, US FCC (2008b), Table 15.

Figure 2.1 Intermodal competition and number of lines or subscribers


connections to make VoIP telephone calls. According to a HarrisInteractive

survey conducted between October 1997 and January 2008, about 15

percent of US adults used VoIP to make telephone calls. The largest US

VoIP supplier was Vonage, which reported about 2.6 million access lines

in August 2008,23 followed by carriers such as Skype, MSN, Yahoo! and

Google. According to InStat, total subscribership for these VoIP services

amounted to about 5 million in 2008.24

In sum, the assumption that ILEC network elements and services are

essential facilities for the supply of retail telecommunications services

is unlikely to hold. Retail customers can buy services from wireline and

intermodal providers, and facilities that a potential competitor requires

to be a wireline CLEC and which must be purchased from on ILEC are

nonetheless not required for some competitors – using some platforms – to

provide competitive retail telecommunications services.

2.4 WHOLESALE PRICE REGULATION: THEORY

In economic theory, assessing the need for ex ante price regulation of a

wholesale market is a very diff erent exercise from its (familiar) counterpart

for retail markets.25 It is only to the extent that a market failure in the

wholesale market impacts prices in the downstream retail markets that

there would be any eff ect on retail customers and thus any economic effi -

ciency consequences to ameliorate with a wholesale regulatory interven-

tion. Said another way, demand for wholesale services is a derived demand,

like the demand for labor services or raw materials. Thus, if the exercise of

market power or the abuse of market dominance for a wholesale service

has no eff ect on downstream retail market prices, customers would not be

harmed, and there would be no likely circumstances under which ex ante

economic regulation would lead to an increase in social welfare or an effi -

ciency justifi cation for economic regulation of the wholesale service.

Conventional regulatory practice ignores this fact when it assesses the

need for wholesale regulation by examining available substitutes for the

wholesale service. In practice, the attempted exercise of market power for

a wholesale service frequently can have no eff ect on retail market prices.

Consider some simple cases where the wholesale service in question is an

essential facility and the ILEC is the only supplier.

2.4.1 Dependent Competition

In Figure 2.2, the only suppliers are wireline carriers: one ILEC and four

dependent CLECs, each of which must purchase wholesale services – say


local loops – from the ILEC. If the CLECs were absent from the retail

market, the ILEC would have a retail monopoly and, if unregulated in

the retail market, would price those services at the profi t- maximizing

monopoly level.

Suppose fi rst that the dependent CLECs eff ectively act as nothing more

than additional distribution channels for the ILEC’s retail service. In

particular, assume that: (1) the CLECs’ incremental costs for the retailing

function are the same as those of the ILEC and (2) the presence of CLECs

in the retail market does not increase total retail demand. If the ILEC

and dependent CLECs pursued a cooperative strategy, their combined

profi ts would be the same as the monopoly profi ts, and the joint- profi t-

maximizing price would be the same as the monopoly price.

In the absence of collusion or wholesale regulation, the ILEC’s profi t-

maximizing wholesale price is the avoided- cost (or effi cient component)

price:

PW = PR – [ICR – ICW] (2.1)

where (PW, PR) and (ICR, ICW) represent the wholesale and retail prices and

incremental costs, respectively.26 To see this formally, it can be shown that

the ILEC’s profi t- maximizing wholesale price is a Ramsey- price markup

of the avoided- cost wholesale price:27

P*W = [hw / (1+ hw)] × [PR – (ICR – ICW)] (2.2)

ILEC

Retail telecommunications services

CLEC 4CLEC 3CLEC 2CLEC 1

Wholesale telecommunicationsservices

Figure 2.2 Dependent competition: monopoly ILEC and oligopoly CLEC


where hw is the wholesale price elasticity of demand. At PW, wholesale

demand is infi nitely elastic (so that P*W = PW) because no CLEC could pay

a wholesale price higher than PW and compete profi tably against the retail

price PR. Intuitively, PW maximizes ILEC profi ts in a non- cooperative

setting, because PW permits the ILEC to extract all of the joint monopoly

profi ts for itself. Perhaps more intuitively, PW and PR maximize ILEC

profi ts because at those prices, the ILEC is indiff erent between serving

customers through its retail and wholesale channels.

Unfortunately, in the absence of regulation in the wholesale market, the

unregulated retail price would be the monopoly price in this case, and con-

sumers would realize no benefi t from the additional retail competition.

Alternatively, if we suppose that CLECs have lower marketing costs

or that they introduce new and innovative services that expand retail

demand, the joint- profi t- maximizing profi ts for a collusion of the ILEC

and CLECs can be higher than the monopoly profi ts of the ILEC alone.28

And in the absence of wholesale regulation, the ILEC can extract some of

that additional profi t by pricing the wholesale service at P*W, which would

exceed PW because CLECs could compete profi tably against PR at whole-

sale prices greater than PW.29

In either case, regulating ILEC wholesale services could reduce the

retail price below the monopoly level and increase consumer welfare. For

example, if the ILEC’s wholesale service were priced at ICW, there would

be fi ve independent, equally effi cient competitors in the retail market

instead of one. What was previously a monopoly would become a more

competitive fi ve- fi rm oligopoly. In the case of Cournot competition,

wholesale regulation would cause the Lerner index of market power for

retail services to fall by a factor of 5, and the retail price would approach

the competitive market price.30 In theory, regulating the wholesale price

makes sense in this setting because every fi rm in the retail market (except

the ILEC) must pay this price, and in the absence of regulation, the ILEC

can use the wholesale price to force the retail price to the monopoly level

and extract the full amount of monopoly profi ts from its retail and whole-

sale customers.

2.4.2 Intermodal Competition

In Figure 2.3, the ILEC–CLEC nexus is identical but we assume the retail

market is also supplied by fi rms that make no use of the ILEC’s wholesale

services. As before, dependent CLECs must purchase network access

services from the ILEC, but wireless, cable and VoIP suppliers use their

own facilities (or, in the case of VoIP, a broadband connection supplied by

the end- user). Suppose the market for retail telecommunications services


is eff ectively competitive, due to the presence of intermodal competitors.

Then, even though the ILEC is, by assumption, the sole supplier of wire-

line network access, it cannot generate any monopoly profi ts from those

services because its CLEC customers could not profi tably pay an excessive

wholesale price and compete against the competitive retail market price.

The demand for wholesale services is derived from the demand for retail

services, and in this case (unlike Figure 2.2), the retail price is unaff ected

by the price of the wholesale service.

Intuitively, dependent CLECs earn normal profi ts at competitive market

wholesale prices. If the ILEC increases the wholesale price, those CLECs

cannot profi tably continue to compete with cable, wireless and VoIP retail

service, which reduces the demand for the ILEC wholesale service. By

assumption, no monopoly profi ts are earned in the retail market; hence,

a wholesale monopolist cannot extract monopoly profi ts from a subset of

the providers of the retail service.

Formally, this situation is illustrated in the expression (2.2) for the

profi t- maximizing wholesale price and is identical to that in Figure 2.2,

except for the level of the retail price. Here, PR is not controlled by the

ILEC and is equal to the competitive market price. As before, at PW, the

demand for wholesale services is perfectly elastic because no CLEC could

compete at a higher price. Thus the profi t- maximizing wholesale price for

the ILEC is PW, the competitive market retail price less the avoided- cost

discount. Again, at these prices, the ILEC is indiff erent between serving

customers through its retail or wholesale channels.

Thus when the retail market is eff ectively competitive, ex ante price

regulation of the wholesale service is unnecessary because the ILEC is

ILEC

CLEC 4CLEC 3CLEC 2CLEC 1

Wholesaletelecommunications services

Retail telecommunications services

Wireless Cable VoIP

Figure 2.3 Intermodal competition: ILEC and dependent and independent

CLEC


unable to extract monopoly profi ts from its (essential) wholesale service.

In fact, its unregulated, profi t- maximizing wholesale price would be

an incremental- cost- based discount from the competitive retail market

service price. In addition, the ILEC would have no direct incentive to

initiate a price squeeze.31 At its unregulated, profi t- maximizing wholesale

price, the regulated retail price would be consistent with (would just equal)

the relevant price fl oor that defi nes an anticompetitive price squeeze.

2.5 WHOLESALE REGULATION IN PRACTICE

2.5.1 The Level of Competition in the Retail Markets

It is useful to consider three cases.

Case 1: Retail market is eff ectively competitive

First, suppose the downstream market is eff ectively competitive due to

the presence of facilities- based competitors that do not use or require the

ILEC’s facilities. This corresponds to Figure 2.3. ILEC wholesale prices

are constrained by competitively determined retail prices, whether or not

there is a separate market for wholesale facilities. Even though the ILEC

is assumed to be the sole supplier of wholesale wireline services, there is no

reason to subject wholesale services to ex ante economic regulation.

In practice, where current (regulated) wholesale prices are not necessar-

ily set at competitive market levels, deregulation in Case 1 could lead to an

increase in wholesale prices. But such a price increase would not constitute

an exercise of market power, which requires a profi table (signifi cant and

non- transitory) increase in price above the competitive level.32 It would

also not be necessarily anticompetitive: pricing that drives effi cient com-

petitors out of an eff ectively competitive market does not reduce social

welfare and is not anticompetitive in economics unless it leads to the

acquisition or retention of signifi cant market power.

Note in this case that the criteria for deregulation may diff er in prac-

tice between wholesale and retail services. In Figure 2.3, an ILEC with

no wholesale competitors may have no market power in the wholesale

market. ‘Defi ning the relevant market for a wholesale facility’ may not

involve any substitute wholesale services. In a recent case, the Canadian

regulator overlooked this point:

37. With regard to future applications to consider the essentiality of a non- mandated service, the defi nition will read as follows: To be essential, a facility, function, or service must satisfy all of the following conditions: (i) The facility


is required as an input by competitors to provide telecommunications services in a relevant downstream market; (ii) The facility is controlled by a fi rm that possesses upstream market power such that denying access to the facility would likely result in a substantial lessening or prevention of competition in the rel-evant downstream market; and (iii) It is not practical or feasible for competitors to duplicate the functionality of the facility.43. The Commission considers that determining duplicability is comparable to defi ning the relevant market for a wholesale facility in the sense that both exer-cises require the identifi cation of potential substitutes, either through existing or potential alternatives.33

Case 2: Retail market is conditionally competitive

The downstream market is eff ectively competitive but that competition

is due, at least in part, to dependent competitors (that is, to dependent

CLECs that use – or require – ILEC facilities). To distinguish Case 2 from

Case 1, the presence of dependent competitors is assumed to be necessary

for competition, that is, that without these carriers, the ILEC would other-

wise possess signifi cant market power in the retail market. Here the ILEC

and the dependent CLECs combined eff ectively possess market power

in the downstream market, in the sense that a hypothetical coordinated

increase in retail prices above competitive market levels could be profi t-

ably sustained. In the absence of price regulation in either market, the

ILEC could then profi tably hold the wholesale price above competitive

levels and capture all of the potential supracompetitive profi ts.

Even though there is no apparent retail market failure in Case 2, the

fact that the necessary competition comes from competitors that are (by

assumption) dependent on the wholesale service of the ILEC means that

the ILEC wholesale service technically fi ts the defi nition of an essential

facility. However, there is nothing in economic theory or in current expe-

rience in telecommunications that suggests this outcome is anything but

a temporary market disequilibrium that is actively evolving towards full

Case 1 facilities- based competition.

Ironically, a major deterrent in this process would be the classifi cation

of wholesale services as essential and ex ante regulation of their price.

Such regulation would deter investment in facilities by the dependent

CLECs and would place intermodal competitors at an artifi cial competi-

tive disadvantage in the retail market. In addition, requiring ILEC to give

competitors access to their network elements at incremental- cost- based

rates distorts the ILEC’s incentive to invest in new facilities and technolo-

gies: incremental costs do not account for either the risk and uncertainty

in investment in new technology or the option value to the CLEC of using

rather than owning facilities. The net eff ect would harm consumers in two

ways. Wholesale regulation would become a self- fulfi lling prophesy in


which competition would never emerge to replace it. And investment by

CLECs and ILECs in this rapidly- developing high- tech industry would be

artifi cially diminished.

As a result, ex post regulation – market observation tempered by compe-

tition law principles – of wholesale services would better serve telecommu-

nications customers than ex ante price regulation of wholesale services.

Case 3: Retail markets are not competitive

Some downstream markets are not eff ectively competitive, for example,

rural or high- cost areas where the ILEC is the only current supplier of

retail services. However, in these markets, ILEC retail services generally

remain subject to economic regulation, so consumers would derive little

direct benefi t from ex ante wholesale regulation. In particular, regulation

of the wholesale price would be – at best – redundant. An ILEC could not

increase its profi ts by charging wholesale prices in excess of:

PW = PR – [ICR – ICW]

where PR is the regulated retail price because no equally effi cient CLEC

could pay that price and compete profi tably.

Indeed, welfare gains from additional retail competition in these

markets may be problematic. It may be the case that entry in retail markets

would be possible in these areas if there were profi table opportunities to

exploit. For example, regulation may have capped prices in rural areas

below a competitive market level, and were the ILEC to price at or above

the competitive level in these areas, entry would more likely occur. In

general, telecommunications networks are deployed incrementally over

time, targeting the most profi table areas – areas having a high concentra-

tion of likely customers – fi rst because higher teledensity areas have more

customers and lower costs of serving those customers. Rural and high- cost

areas may be targeted in later stages so that the current lack of competitors

may not necessarily refl ect a natural monopoly condition for which ex ante

regulation might be appropriate.

2.5.2 Summary

In current telecommunications markets, retail services are provided over

at least three distinct platforms – traditional narrowband wireline, wireless

and cable – and VoIP suppliers provide service over unaffi liated broadband

access facilities. Where retail markets are competitive and consumers can

choose among multiple platforms, there is no market failure for regulation

to redress. Thus in Case 1, ex ante regulation of wholesale services on one


of those platforms – the wireline platform – cannot improve consumer

welfare and the inherently asymmetric regulation of one platform relative

to others would likely reduce consumer welfare. For Case 3, regulation

of some retail services remains in place and, as discussed, there is no need

for additional price regulation of wholesale services. Effi cient regulation

of wholesale services in Case 2 depends critically upon the facts in the

market. If there is uncertainty as to whether the ILEC wholesale service

is essential, a reasonable policy would be to undertake a market- based

experiment: permit but not require the unbundling of the wholesale service

at regulated prices and allow the market to correct itself from the distor-

tions caused by current overinclusive wholesale regulation. Observe the

course of competition in the retail market to determine ex post whether the

wholesale service is truly essential. Given the ubiquity of facilities- based

competition (cable, wireless and VoIP providers) in telecommunications

markets, it would be surprising to fi nd – even in markets dominated today

by an ILEC and dependent CLECs – that the retail market was inherently

dependent on access to the ILEC’s wireline facilities.

That experiment has been undertaken in the US, where unbundling

requirements on ILECs have been relaxed and, in particular, the com-

bined loop and switch (UNE- P) mandatory off ering has been replaced

with negotiated commercial contracts. All indications suggest that local

exchange competition remains robust, as ILEC customer counts have

fallen, while customer counts of cable companies, wireless carriers and

VoIP suppliers have grown.

2.5.3 Ex post Regulation of Anticompetitive Conduct

Generally speaking, throughout the economy, anticompetitive conduct

and abuse of dominant position are regulated and controlled on an ex

post rather than ex ante basis through the enforcement of competition

laws. Advantages of ex post regulation of anticompetitive conduct with

respect to essential facilities include the reduced risk of economic distor-

tions (for example, changing the basis of negotiation for use of facilities)

and the effi ciency gain from intruding only in cases where anticompetitive

behavior actually occurs. Thus, across many diff erent jurisdictions, ex

post regulation of provision and pricing of essential facilities is the norm

for other industries, and the network externalities and network eff ects that

distinguish telecommunications from many other industries do not justify

industry- specifi c ex ante economic regulation.

For telecommunications services, what is the potential eff ect in retail

telecommunications markets from the anticompetitive denial or pricing

of access to essential wholesale services that ex post regulation would


be expected to control? In economics, such behavior is termed a price

or margin squeeze, where a vertically integrated fi rm (for example, an

ILEC) prices its wholesale service at a level that prevents an otherwise

effi cient dependent competitor from competing against the price of the

ILEC’s retail service.34 A fi rm that sets a wholesale price for an essential

facility that (together with the retail market price) entails an anticom-

petitive price squeeze necessarily sacrifi ces profi ts, at least in the short

run because it loses profi ts on every retail service it sells. A component

of the cost to an ILEC of supplying the retail service is its opportunity

cost – the contribution (wholesale price less wholesale incremental cost)

from the wholesale service – that is forgone when the ILEC serves the

retail customer.

Currently in telecommunications markets, there is little likelihood for

such behavior to happen because the conditions necessary for a price

squeeze to be profi table do not generally hold. Such a sacrifi ce of short-

run profi ts is unlikely to be profi table in the long run in telecommunica-

tions markets because of the diffi culty of recoupment – of eliminating

competitors and then raising prices above a competitive market level.

Characteristics of telecommunications markets that make recoupment

diffi cult include: (1) the absence of retail barriers to entry; (2) the presence

of competitors (for example, cable and wireless companies) that do not

depend on the putative essential facility, that have sunk costs and that

provide retail services in markets other than retail telecommunications; (3)

the pace of technological change; and (4) the rates of growth of markets

for retail telecommunications services. Like predatory pricing, denial or

anticompetitive pricing of an essential facility is a logical possibility in

telecommunications markets but is suffi ciently unlikely that it makes little

sense to regulate access and pricing of essential facilities on an ex ante

basis.

Finally, the determination that an ILEC facility is essential and must be

provided on an unbundled basis ought not to be an ex ante exercise. At

market- determined prices, some ILEC services may generate substitutes

that are not viable at regulated prices or prices determined by mandatory

arbitration. A less intrusive and more market- consistent approach would

permit market forces to set wholesale prices and would classify a service or

facility as essential only on an ex post basis after complaint and a fi nding

by the regulatory or competition authority. Only after a fi nding of essenti-

ality would the commercially negotiated prices (and other terms and con-

ditions) for those facilities determined ex post to be essential be required.

And only in the case that the parties could not agree to prices, terms and

conditions would regulatory or competition authorities’ step in on an ex

post basis to resolve the diff erences.35


2.6 CONCLUSIONS

In many telecommunications markets, ex ante economic regulation of

both wholesale and retail services is generally unwarranted, mutually

inconsistent and rife with ineffi cient, unintended consequences. Where

retail markets are eff ectively competitive due to intermodal competition,

regulation of wholesale services serves no function and is unnecessary even

though there may be no substitutes for ILEC wholesale services. Even

where retail markets are not eff ectively competitive, regulation of whole-

sale services may not be warranted: substitutes for the ILEC’s wholesale

facilities may be forthcoming at competitive market rather than regulated

prices. In both cases, the fact that wholesale regulation cannot be con-

sistent across intermodal platforms is a well- known recipe for ineffi cient

outcomes.36

Ex post regulation of wholesale services through application of antitrust

and competition law can be applied at parity across these diff erent plat-

forms and would avoid many of these problems of ex ante regulation.

NOTES

1. The term ‘CLEC’ is often used to refer to non- incumbent local exchange carriers generi-cally, not necessarily restricted to carriers using traditional wireline technology. Thus, in many states in the US, cable companies are considered CLECs for some regulatory purposes. For simplicity, I use the term ‘CLEC’ to mean ‘wireline CLEC’ here, and use ‘intermodal carrier’ to include non- wireline competitors that may or may not be CLECs.

2. Note that VoIP suppliers, unlike cable companies and wireless carriers, do not generally provide their own network infrastructure but rather supply their services as applica-tions carried over broadband facilities supplied by another carrier.

3. See, for example, US FCC 1996. 4. Propounded by Martin Cave (2006), this ‘ladder of investment’ theory has been widely

discussed as an underpinning to regulatory policy at the EU and elsewhere. 5. In the US, Section 251(3)(2)(B) of the Telecommunications Act of 1996 requires ILECs

to unbundle network elements and provide them at cost- based prices whenever com-petitors would be ‘impaired’ without access to them.

6. Countries requiring some form of local loop unbundling include the US, Canada, Australia, New Zealand, and the European Union member countries, Hong Kong and South Africa.

7. See, for example, EU Access (European Parliament and Council of the EU, 2002a). 8. See, for example, in Canada, Telecom Decision CRTC 2008- 17 (CRTC, 2008); or

in Europe, EU SMP Guidelines (European Parliament and the Council of the EU, 2002b).

9. The unbundled network element platform (UNE- P) was a combination of unbundled local loop and switching which was essentially a form of very cheap resale. Total element long- run incremental cost (TELRIC) is the cost standard in the US for unbun-dled network elements. Depending upon the regulators’ interpretation, this standard resembles the economist’s total service long- run incremental cost.


10. See for example, J.R. Abel and V. Watkind- Davis (2000), pp. 14–15.11. For example, cheap wireline access (UNE- P in the US) gave wireline CLECs an

uneconomic advantage to compete successfully for a period against cable and wireless suppliers.

12. See, for example, US Department of Justice (2008).13. US FCC (2006).14. NCTA (n.d.).15. Comcast Press Release – (2008).16. US FCC Twelfth CMRS Report (US FCC, 2008a).17. ‘25 percent of wireless users were business customers, with the remaining 75 percent

being ordinary consumers.’ (US FCC, 2008a), footnote 633.18. There were approximately 94 million residential ILEC plus CLEC switched access lines

in December 2007. US FCC, 2009 Local Competition Report, Table 2.19. All of the growth in wireless- only households may not refl ect substitution because there

has also been an increase in the number of households, some of which may never have had a wireline phone. However, the relevant numbers are small. About 1.5 million households were added on average over the last fi ve years according to US Census data. Assuming that 21 percent of these households (that is, the average percentage of ‘wireless- only’ 18- 29- year- old adults) decided not to install a wireline phone implies that only about 320 000 new households per year have never purchased wireline access. Over that same period, the number of wireless- only households has grown by 4.2 million per year. Thus, the number of households that substituted wireless for wireline service has averaged about 3.8 million per year or 92 percent of the total gain in the wireless- only count.

20. US FCC (2008b).21. Ibid., Table 14.22. Ibid., Table 14.23. Press release at http://www.marketwatch.com/news/story/vonage- cuts- quarterly- loss-

only/story.aspx?guid=%7BFE7B3F24%2DFAC6%2D4300%2D8867%2D76728F3DB6B3%7D&dateid=39667.3494664699- 935212285.

24. http://www.instat.com/promos/08/dl/IN0804301WWI_as78sa.PDF .25. For a retail market, economists would generally advise forbearance when no carrier

possessed market power as indicated by such structural market characteristics as the presence of alternative suppliers and low barriers to entry and expansion. See for example, the EU SMP Guidelines (European Parliament and Council of the EU, 2002b, p. 5).

26. Note that ICR is the incremental cost of the entire retail service. ICR – ICW can be thought of as the ILEC’s incremental cost of the retailing function, assuming that the ILEC’s cost of using the wholesale service itself is the same as the cost of providing that service to CLECs.

27. See Taylor (2008), p. 18.28. Intuitively, CLEC demand is higher at the monopoly price and CLEC incremental costs

are lower.29. This is an interesting special case because the ILEC fi nds it profi table to set a wholesale

price above its avoided- cost discount. In economics, such wholesale and retail prices constitute a vertical price squeeze, and an open question in antitrust enforcement and competition policy is whether such prices necessarily entail a reduction in the ILEC’s profi ts.

30. See, for example, Carlton and Perloff (2005), pp. 194–7.31. There may still be an indirect incentive to price squeeze, analogous to predatory

pricing, in which current profi ts are sacrifi ced in the expectation of driving rivals from the market and recouping lost profi ts in the future. While a logical possibility, the pre-conditions for such a strategy to be successful do not often hold in telecommunications markets. In addition, as above, if CLECs were more effi cient than the ILEC in provid-ing the retail function, the profi t- maximizing wholesale price could exceed PW because


CLECs could profi tably compete – at least as well as the ILEC – at a wholesale price above PW, so that wholesale demand would still have fi nite elasticity at PW and P*

W would exceed PW.

32. As defi ned by the US DOJ Merger Guidelines, ‘market power is the ability profi t-ably to maintain prices above competitive levels for a signifi cant period of time’ (US Department of Justice and Federal Trade Commission, 1992, § 0.1).

33. Telecom Decision (CRTC, 2008).34. For this discussion, denial of access is envisioned (constructively) as simply pricing

access too high to permit any effi cient fi rm to compete.35. It does not follow that the price and other terms and conditions should be set on an ex

ante basis and tariff ed. Costs and market conditions for essential facilities can diff er signifi cantly across potential customers, so that economic effi ciency would be sacrifi ced by an ex ante regulatory determination of prices and terms and conditions irrespective of market circumstances.

36. The classic example of these distortions is US regulation of surface transportation in the 1950–80 time period and the resulting ineffi cient mix of railroad, truck and barge infrastructure investment and, ultimately, traffi c.

REFERENCES

Abel, J.R. and V. Watkins- Davis (2000). Geographic deaveraging of wholesale prices for local telephone service in the United States: some guidelines for state commissions. National Regulatory Research Institute. April.

Canadian Radio- television and Telecommunications Commission (CRTC) (2008). Canada Telecom Decision (CRTC, 2008- 17), 3 March.

Carlton, Dennis W. and Jeff rey M. Perloff (2005). Modern Industrial Organization, 4th edn (pp. 194–7). Boston, MA: Pearson Education.

Cave, M. (2006). Encouraging infrastructure competition via the ladder of invest-ment. Telecommunications Policy, 30, 223–37.

Comcast (2008). Press release: third quarter 2008 results. Table 6. Retrieved from http://media.corporateir.net/media_fi les/irol/11/118591/Earnings_3Q08/3Q08PR.pdf.

European Parliament and the Council of the European Union (2002a). Directive 2002/19/Ec on access to, and interconnection of, electronic communications networks and associated facilities. EU Access Directive. Retrieved from http://www.ictregulationtoolkit.org/en/Pusblication.1496.html. 7 March.

European Parliament and the Council of the European Union (2002b). European Commission Guidelines on Market Analysis and the Assessment of Signifi cant Market Power under the Community Regulatory Framework for Electronic Communications Networks and Services (2002/C 165/03) (‘EU SMP Guidelines’). Retrieved from http://www.ictregulationtoolkit.org//en/Publication.2589.html.

National Cable and Telecommunications Association (NCTA) (n.d.). Industry data. Retrieved from http://www.ncta.com/Statistic/Statistic/Statistics.aspx.

Taylor, W. (2008). Intermodal telecommunications competition: implications for regulation. International Telecommunications Society. Retrieved June 2008 from http://www.canavents.com/its2008/abstracts/27.pdf.

Telecommunications Act of 1996, Pub. L. 104- 104, 110 Stat. 56 (1996). Section 251(3)(2)(B).


US Department of Justice (DOJ) and Federal Trade Commission (FTC) (1992). Horizontal Merger Guidelines, April (‘DOJ Merger Guidelines’) § 0.1.

US Department of Justice (DOJ) (2008). Voice, video and broadband: the chang-ing competitive landscape and its impact on consumers. November. Retrieved from http://www.usdoj.gov/atr/public/reports/239284.pdf.

US Federal Communications Commission (FCC) (1996). Implementation of the Local Competition Provisions of the Telecommunications Act of 1996, CC Docket Nos. 96- 98, 95- 185, First Report and Order (‘Local Competition Order’) 11 FCC Rcd 15499, 231–232.

US Federal Communications Commission (FCC) (2006). In the Matter of Annual Assessment of the Status of Competition in the Market for the Delivery of Video Programming (MB Docket No. 05- 255). Twelfth Annual Report. Released: 3 March, Table 1.

US Federal Communications Commission (FCC) (2007). Local Telephone Competition: Status as of June 30, 2006 (‘2007 Local Competition Report’), released January 2007, Tables 1, 14.

US Federal Communications Commission (FCC) (2008a). Annual Report and analysis of Competitive Market Conditions with Respect to Commercial Mobile Services, Twelfth Report (‘Twelfth CMRS Report’), FCC 08- 28, released 4 February, Table 1.

US Federal Communications Commission (FCC) (2008b). High- Speed Services for Internet Access: Status as of June 30, 2007 (‘2008 High- Speed Internet Report’) released March, Table 15.

US Federal Communications Commission (FCC) (2009). Local Telephone Competition: Status as of June 30, 2008 (‘2009 Local Competition Report’), released July 2009, Tables 1, 2, 14.

63

3. Access regulation versus infrastructure investment: important lessons from Australia1

Martyn Taylor

3.1 INTRODUCTION

A key focus of increasing concern in telecommunications regulatory policy

is the impact of access regulation on the long- term incentives for new

infrastructure investment. This chapter analyses the theory and practice

behind the regulation versus investment debate and makes a series of key

policy recommendations. Importantly, many jurisdictions have not yet

struck an appropriate balance, particularly in circumstances where sub-

stantial investment in next generation networks is now required.

3.2 DETERMINANTS OF TELECOMS INFRASTRUCTURE INVESTMENT

The Internet is one of the most signifi cant innovations of the twentieth

century and is profoundly shaping modern society in the twenty- fi rst

century. Governments around the world are increasingly recognizing the

value of Internet access, including via positive externalities that spill over

into economic growth. A common policy vision among advanced industri-

alized nations is ubiquitous broadband Internet access.

A key constraint to the supply of high- speed broadband access into

the home in most countries is legacy wireline ‘last mile’ customer access

network infrastructure. With the exception of new housing developments,

such infrastructure in most nations typically comprises either a twisted

copper pair, often many decades old and prone to quality issues; or, less

commonly, a separate coaxial cable, normally as part of a hybrid fi ber

coaxial network initially deployed to provide cable television services.

Broadband Internet access is normally provided over such infrastruc-

ture via xDSL (digital subscriber line) and DOCSIS (Data Over Cable


Service Interface Specifi cation) technologies, respectively. However, both

technologies have inherent limitations borne from the quality of the under-

lying infrastructure. The policy vision of ubiquitous broadband Internet

access is therefore underpinned by the rollout of so- called next generation

networks (NGNs) that upgrade legacy infrastructure by extending high-

capacity fi ber optic cable towards the consumer.

3.2.1 Realizing the Vision: the Economics of Infrastructure Investment

The rollout of NGN infrastructure involves signifi cant labor and capital

sunk costs. The investment is essentially irreversible: once deployed, fi ber

cannot easily be redeployed. As the fi rm’s ability to redeploy capital is

limited, its potential losses from poor investment decisions are commen-

surately greater.

The investment also involves economies of scope and scale and network

eff ects that make it optimal to roll out the network rapidly on a holistic

rather than piecemeal basis. As a consequence, the investment is ‘lumpy’

and most costs are borne immediately rather than spread over the network

life. In contrast, revenues are normally recovered over the substantial life-

time of the network assets, creating a signifi cant timing mismatch between

outgoing cost cashfl ows and incoming revenue cashfl ows. Financing

arrangements must address this mismatch. Given such issues and the

magnitude of the costs involved, NGN investment decisions are serious

matters for industry participants.

Generally, a fi rm should only invest in telecommunications infrastruc-

ture if it can achieve aggregate revenue cashfl ows over the project life that

exceeds its aggregate cost cashfl ows, adjusting for timing, fi nancing and

opportunity costs. Under a standard net present value (NPV) analysis,

this is assessed by discounting the incremental net ungeared after- tax cash

fl ows of the project at an appropriate weighted average cost of capital

(WACC). The cashfl ows are probability- weighted to adjust for project-

specifi c risks.

Any investments achieving an average return greater than the WACC

create value and are represented by a positive NPV. Conversely, any

investments achieving an average return below the WACC are value

destructive and are represented by a negative NPV. Most fi rms choose

between investment projects based on their NPV ranking.

The WACC is the fi nancing cost of the project, calculated as the

weighted average of the cost of debt and equity. The precise weighting is

determined by the capital structure of the project: debt increases leverage

while providing a benefi t via tax- deductible interest payments. The cost

of debt depends on the credit rating of the borrower. The cost of equity is

Access regulation vs infrastructure investment 65

usually determined by the capital asset pricing model (CAPM) as the risk-

free interest rate, plus a beta- adjusted market risk premium.

3.2.2 The Impact of Risk on Project Decision- Making

Bearing this decision- making framework in mind, the concept of ‘risk’

aff ects investment decision- making in two principal ways:

Symmetric risks: risks with similar upside and downside impacts (for ●

example, changes in interest rates) increase the volatility of project

cashfl ows. Such generic market risks are considered ‘undiversifi able’

for equity investors under modern portfolio theory so are factored

into the cost of equity, leading to a reduced NPV. The project must

therefore earn a higher rate of return before it proceeds relative to

competing projects.

Asymmetric risks: risks with asymmetric upside and downside ●

impacts (for example, price regulation) are considered ‘project-

specifi c’ and therefore ‘diversifi able’ for equity investors under

modern portfolio theory so are not factored into the cost of equity.

Rather, such risks are weighted directly into the project cashfl ows

and hence similarly reduce NPV. Again, the project would need to

earn a higher rate of return before it proceeds relative to competing

projects.

However, the level of return obtainable from telecommunications infra-

structure is constrained by market conditions. Project revenue is a func-

tion of price and the volume of services supplied. The price is constrained

by consumers’ willingness to pay, competition from substitutable services

(for example, wireless broadband), and regulatory constraints. Volume is

determined by price under a market demand curve.

An NGN investment will not therefore proceed unless and until a fi rm

has reasonable certainty (in light of forecast long- term market conditions

and project- specifi c risks) that it will earn a return that exceeds its cost of

its capital over the long- term project life, including a market risk- adjusted

return on that capital.

3.3 THE IMPACT OF REGULATION ON INVESTMENT

There are a variety of circumstances beyond generic competition law

in which policy- makers have intervened to regulate markets in the


public interest. Sectoral telecommunications regulation is an important

example.

However, well- intended regulation can have unintended adverse eff ects.

In essence, regulation can harm investment incentives by increasing the

risk faced by investors without providing scope for off setting higher

returns. To illustrate this eff ect it is useful to consider the impact on tel-

ecommunications infrastructure investment of network unbundling and

cost- based access regulation.

3.3.1 Excessive Regulation as a Deterrent to Infrastructure Investment

Mandated network unbundling has long been justifi ed as promoting alloc-

ative effi ciency by increasing the level of competition to which incumbent

network owners are subject. Market entrants can selectively bypass exist-

ing infrastructure via their own network elements while obtaining exclu-

sive use rights to remaining network elements at cost- based access prices.

Unbundling is intended to lower barriers to market entry while enabling

competitors to aggregate suffi cient customers to enable them to fi nance a

more extensive infrastructure rollout.

While unbundling undeniably stimulates short- term competition, this

may often be at the expense of long- term investment incentives. The inter-

national trend to balance this trade- off has been to roll back unbundling

obligations as competition develops. The trend towards such regulatory

‘forbearance’ is exemplifi ed by the following comment by the United

States Federal Communications Commission (FCC) in 2003 in conjunc-

tion with its subsequent decisions:

While unbundling can serve to bring competition to markets faster than it might otherwise develop, we are very aware that excessive network bundling requirements tend to undermine the incentives of both incumbent LECs [local exchange carriers] and new entrants to invest in new facilities and deploy new technology. (FCC, 2003a, para 3)

The rationale for such forbearance is well explained in the literature.

Professor Sidak commented in a 2006 journal article that the FCC imple-

mentation of unbundling led to: ‘tens of billions of dollars of investment

fl owing into business models that were neither particularly innovative nor

sustainable in the absence of regulatory distortions in their favour. That

distortion of investment represented a staggering destruction of wealth’

(Sidak, 2006, p. 55).

Professors Gayle and Weisman concluded in a 2007 article that United

States competitive local exchange carriers ‘proceeded to lose billions of

dollars when the FCC’s introduction of [unbundling] along with artifi cially


low prices for network elements resulted in what was ultra- free- entry’,

and, ‘policies that reward imitation rather than innovation will attract

those market entrants adept at innovation, predominantly arbitragers,

while driving away genuine innovators’ (Gayle and Weisman, 2007, p.

321). Importantly, the genesis of the concerns with unbundling involves

the price at which access is provided to unbundled network elements.

3.3.2 Criticisms of Cost- Based Incentive Pricing: TELRIC and its

Variants

The impact of unbundling on investment is a function of the price at which

access is provided to unbundled infrastructure. The selection of that price

involves a balancing exercise between promoting allocative and dynamic

effi ciency:

If the access price is set too low, a fi rm will have a reduced incen- ●

tive to invest in new infrastructure: hence dynamic effi ciency will be

impeded.

If the access price is set too high, a fi rm will face less competition and ●

will have greater market power to set prices at levels that maximize

its profi ts: hence allocative effi ciency will be impeded.

Most nations have sought to balance allocative and dynamic effi ciency

by adopting so- called ‘incentive regulation’. Such regulation is intended to

mimic competitive markets by constraining prices at a level consistent with

that which would otherwise be imposed by competition.

The United States has historically used a form of incentive regulation

known as ‘total element long- run incremental cost’ (TELRIC). Other

jurisdictions have used variants of this, including total service long- run

incremental cost (TSLRIC). TELRIC and its variants determine prices

based on the incremental costs faced by an effi cient cost- minimizing fi rm

with an optimally confi gured network that uses the best available current

technology. The network elements in the hypothetical network are nor-

mally assumed to be located at the same geographic nodes as the regulated

fi rm’s existing network.

In this manner, the price of access is determined not by the regulated

fi rm’s need to achieve a return on its historic capital costs, but rather by

a forward- looking model in which the only costs recoverable are those

incurred by a hypothetical effi cient fi rm. The policy rationale for this

approach is that a network provider in a competitive market would be

constrained in its pricing by the price of its most effi cient competitor.

Under the TELRIC philosophy, that competitor is assumed to have


the most effi cient network and aggressively undercut the incumbent’s

pricing.

In theory, if the TELRIC approach were applied accurately, any invest-

ment in network infrastructure would realize an NPV near zero. The fi rm

would not earn above the project WACC. Professor Pindyck explained

this in a statutory declaration in the following terms (FCC, 2003a):

The TELRIC pricing methodology relies on the simple Net Present Value (NPV) investment rule. The NPV rule states that a fi rm should invest in a project if the sum of the discounted cash fl ows (the NPV) from the project is positive. TELRIC is designed to produce prices that price an ILEC with a competitive return, i.e., with no profi ts in excess of those that would arise in a competitive market. In other words, under TELRIC the expected NPV of the included costs at any given discount rate is zero. The theory behind the TELRIC methodology is that, if the NPV was greater than zero, additional fi rms would enter until excess profi ts were driven to zero.

TELRIC pricing creates a clear disincentive towards new infrastruc-

ture investment. By investing in a project subject to TELRIC pricing, the

fi rm is exposed to little upside to create shareholder value. Rather, the

fi rm is exposed to signifi cant downside that could destroy shareholder

value. In such circumstances, a rational fi rm would choose to invest its

money in alternative projects that do create shareholder value via a posi-

tive NPV.

Beyond this immediate NPV = 0 issue, there are a range of other dif-

fi culties with the application of TELRIC pricing. These diffi culties create

signifi cant asymmetric downside risk and therefore further reduce the

project NPV.

Firstly, rapid technological innovation and regular cost reviews create

diffi culties. Continued innovation usually results in the price of modern

telecommunications technology declining over time for a given function-

ality. If TELRIC pricing is regularly reassessed, such declining prices will

mean that a fi rm will never recover the initial cost of its investment. The

fi rm will only be permitted to recover its historic costs to the level of the

best available technology at the time of the review.

Generally, the shorter the TELRIC review period and the faster the rate

of technological innovation, the greater the risk to the fi rm that it will not

recover its costs. Professors Mandy and Sharkey (2003) have identifi ed

that the magnitude of this loss is considerable: if asset prices fall 11 percent

per year, TELRIC pricing leads to an under- recovery of around 50 percent

(Mandy and Sharkey, 2003, p. 437).

Secondly, the regulator always has the benefi t of perfect hindsight. In

investment decision- making, the fi rm will have imperfect information


regarding future demand. The higher standard deviation of such forecasts

may lead to a fi rm to overdimension its network to address high- demand

scenarios.

In contrast, the regulator has perfect hindsight. At the time it makes its

regulatory decisions, it will know actual traffi c patterns and can ascertain

the optimal network to meet those patterns. This diff erence in information

quality means that regulators are often skeptical of network dimensioning

and have a propensity to disallow costs.

Thirdly, competitors share in rewards but not risks. In making an

investment, the fi rm is exposed to the market risk of incorrect demand

forecasts. If the network is a failure, the fi rm alone will bear the cost of

that failure. However, if the investment is a success, access regulation

will mean the fi rm is subject to access requests from competitors (‘access

seekers’) wishing to share in that success at cost- based access prices. In

this manner, regulation exposes the fi rm to an asymmetric project return

profi le with an unlimited downside risk but truncated upside benefi t. Such

an asymmetric profi le will negatively skew the NPV calculation and is a

disincentive to investment.

Furthermore, the disincentive to invest is even more acute when con-

sidered under ‘real option’ theory. The regulated fi rm must make a long-

term irreversible investment, so is committed to its substantial investment

for the network life. The fi rm foregoes the value of an option to wait for

lower- cost technology and is exposed to the downside risk that forecast

demand does not materialize. In contrast, the access seeker retains fl ex-

ibility and has a valuable option to enter or exit the industry at will, or to

roll out its own infrastructure with a cheaper technology should demand

forecasts be correct.

Fourthly, free- riding and disincentives for competitors to invest cause

concerns. Not only will TELRIC pricing create a disincentive for the

regulated fi rm to invest, but the pricing will also create a signifi cant dis-

incentive for access seekers to invest. Access seekers will rank the NPV of

investment options when making their investment decisions. A rational

access seeker will prefer the investment delivering the highest NPV. It will

not incur the signifi cant capital cost and long- term risk involved in rolling

out infrastructure, if it can obtain immediate access to such infrastructure

at a TELRIC price consistent with that of the best technology available.

The access seeker would prefer to free- ride on the investment by the regu-

lated fi rm rather than rolling out its own competing infrastructure.

Fifth, and fi nally, is regulatory error. Regulators are not perfect and

may exhibit a natural bias in favor of the market entrant or may be prone

to error. The scope for regulatory error is particularly high in the context

of highly complex telecommunications pricing models given the sheer


range of interacting assumptions, their limited verifi ability and associated

measurement complexity.

Given these concerns, it is not surprising that Professors Crandall and

Sidak commented in a June 2007 review of the literature that: ‘the net

eff ect of mandatory unbundling on dynamic effi ciency (investment by both

entrants and incumbents) is therefore unequivocally negative’ (Crandall

and Sidak, 2007, p. 10).

3.3.3 Conclusions on the Impact of Regulation on Investment Incentives

For existing (legacy) telecommunications infrastructure, there is little risk

to regulators that such infrastructure will be removed if excessively low

access prices are imposed. The cost of such infrastructure is sunk. For

existing infrastructure owners, low access prices will largely transfer value

from the infrastructure owner’s shareholders to the shareholders of access

seekers. Any impact on investment is limited to ongoing maintenance of

the network.

However, the situation is very diff erent for new (greenfi eld) infrastruc-

ture. If excessively low access prices are imposed, the infrastructure owner

has the ability to avoid incurring the sunk cost by not undertaking the

investment. In this manner, low access prices will likely result in the infra-

structure not being built. Low access pricing therefore has a direct and

disproportionate impact on the rollout of new infrastructure.

Bearing this important distinction in mind, international best practice

in telecommunications regulation has historically favored a high degree of

network unbundling of legacy network infrastructure with the associated

application of cost- based access regulation. While such regulation has

promoted market entry and services- based competition, it has impeded

infrastructure investment and longer- term facilities- based competition.

Specifi cally, it has deterred investment in greenfi eld infrastructure and the

upgrading of legacy networks. Given such issues, the optimal regulatory

approach should be to wind back such regulation as competition develops

so as to preserve long- term network investment incentives.

The unbundling experience demonstrates that care is required in the

application of telecommunications price regulation to ensure that an

optimal balance is achieved between allocative effi ciency and dynamic

effi ciency. While TELRIC pricing is intended to achieve this balance, the

current application of TELRIC has seriously impeded dynamic effi ciency

by deterring infrastructure investment. Serious policy concerns there-

fore arise regarding the continued appropriateness of applying TELRIC

pricing to greenfi eld telecommunications infrastructure investments or

network upgrades.


3.4 AN AUSTRALIAN CASE STUDY IN EXCESSIVE REGULATION

As a geographically vast country with a highly dispersed population, located

at a great distance from its trading partners, Australia depends heavily on

the quality, effi ciency and innovativeness of its telecommunications system.

However, attracting suffi cient funds for investment in Australian fi xed tele-

communications infrastructure has been a signifi cant challenge.

3.4.1 Adoption of a Negotiate–Arbitrate Model Underpinned by TSLRIC

Pricing

Australia was one of the earliest countries to liberalize its telecommunica-

tions sector, commencing an interim phase of liberalization from 1991.

Australia subsequently implemented full liberalization under legislative

reforms in 1997. A sectoral telecommunications access regime known as

‘Part XIC’ was enacted into the Trade Practices Act 1974 (Cth) based on

a negotiate–arbitrate model.

Part XIC enables the Australian Competition and Consumer

Commission (ACCC) to declare various wholesale telecommunications

infrastructure and carriage services after a formal public inquiry on the

basis that the declaration is in the ‘long- term interests of end- users’.

Following declaration, various standard access obligations apply that

require an access provider to supply a declared service to access seekers

on request on a non- discriminatory basis. There are currently 9 declared

services in Australia and they include unbundled network elements.

A consequence of declaration under Part XIC is that, if the access

provider and access seeker cannot agree on any aspect of the supply of a

declared service, either party may notify the ACCC and seek arbitration.

Most access arbitrations are lodged by access seekers and relate to the

price at which the declared service is supplied. The ACCC has published

non- binding pricing principles for each declared service to guide its arbi-

tral determinations. The favored pricing methodology by the ACCC has

been TSLRIC pricing, a variant of the TELRIC pricing methodology.

However, the manner of application of TSLRIC pricing by the ACCC

has created signifi cant concern in Australia. TSLRIC pricing has deterred

investment in infrastructure that could be subject to actual or potential

Part XIC declaration. In this manner, investment fl ows have been dis-

torted away from regulated (and potentially regulated) services and infra-

structure towards unregulated services and infrastructure.

Furthermore, not only has the regulatory regime created a disincentive

for investment; but the absence of greenfi eld investment has been used as


political justifi cation to perpetuate and continually strengthen the regula-

tory regime. This regulation has further exacerbated the risk of investing

in greenfi eld infrastructure, thereby feeding a vicious cycle.

In this manner, Australia has been caught in an upwards spiral

of increasing sectoral telecommunications regulation. This spiral has

occurred even though the level of competition in Australian telecommu-

nications markets has substantially increased. According to international

best practice and economic theory, the level of sectoral regulation should

rather have been rolled back as competition developed.

3.4.2 Refi nements to the Regulatory Regime to Promote Greater

Investment

The potential adverse impact of the Part XIC access regime on investment

was recognized in Australia as early as 2001, only four years after the

enactment of Part XIC. In a comprehensive review of the telecommunica-

tions regulatory regime in 2001, the Australian Productivity Commission

commented:

Mandated access still presents formidable regulatory risks to investors. Telecommunications technology and markets are rapidly moving and very risky . . . For a carrier making a new investment; the risk of future declara-tion – with regulated access prices – may lead to the delay or termination of the planned investment. This is because access regimes may truncate the returns from risky investments.

The Productivity Commission made a series of recommendations to

promote investment, principally the adoption of an ‘access holiday’

regime.

In December 2002, Australia therefore enacted access holiday provi-

sions into Part XIC, known as ‘anticipatory individual exemptions’. A

carrier could apply to the ACCC for an exemption if it would promote

the long- term interests of end- users, assessed on the basis of promoting

competition and encouraging the economically effi cient use of, and invest-

ment in, infrastructure. The ACCC could grant the exemption subject to

any conditions. The explanatory memorandum to the relevant legislation

indicated that the purpose of anticipatory individual exemptions was to

provide certainty for potential investors in telecommunication infrastruc-

ture and services in relation to future access.

However, the exemption mechanism met immediate diffi culties the fi rst

time it was applied. The suppliers of the Australian analogue (FOXTEL)

cable television service sought an anticipatory individual exemption to

underpin their upgrade of network assets and equipment to support digital


cable television. The ACCC granted the exemption subject to various con-

ditions. The parties proceeded with the upgrade on the basis of the exemp-

tion. However, the ACCC’s decision was subsequently appealed.

In hearing the appeal, the Australian Competition Tribunal noted

that the parties had proceeded with the rollout of digital cable television

before the appeal had been determined. The parties had also made a range

of statements in the public domain regarding their intention to roll out

digital television without qualifying these by the need for the appeal to be

determined. The Tribunal therefore held that the network upgrade would

have proceeded regardless of the exemption; the exemption was unneces-

sary for the investment to proceed.

Even though the digital upgrade had already occurred on the basis of

the exemption, the Tribunal overturned the exemption, leaving the parties

exposed to the risk of access regulation. Ironically, the Productivity

Commission had used the rollout of digital cable television as a key

example of the need for the ‘access holiday’ mechanism in the fi rst place.

Following the Tribunal decision, anticipatory individual exemptions

would only be eff ective in Australia if an access provider delayed its invest-

ment until the exemption had been granted and any appeals of that decision

were resolved. This could take at least two years and potentially signifi -

cantly longer given the potential for further appeals to higher courts.

Not surprisingly, no further applications were made to the ACCC

for anticipatory exemptions. Ironically, the mechanism intended by the

Australian government to facilitate investment had the unfortunate con-

verse impact of further impeding it.

3.4.3 The Saga of Australia’s Fiber to the Node Network

Issues with the application of Part XIC to new infrastructure invest-

ments came to a head between Australia’s incumbent telecommunications

carrier, Telstra Corporation Limited, and the Australian government

during 2006 and 2007 in the context of the proposed upgrade of Telstra’s

network to implement fi ber to the node (FTTN).

Given the perceived diffi culties with Part XIC, Telstra indicated that it

was not intending to seek an anticipatory individual exemption from the

ACCC. Rather, Telstra sought to enter into negotiations directly with the

government with a view to obtaining direct legislative relief from the Part

XIC access regime. Negotiations between Telstra and the government

occurred throughout much of 2006. Telstra’s requested the government to

enact legislation that disapplied Part XIC to Telstra’s network upgrade,

eff ectively to provide a long- term access holiday by way of specifi c

legislation.


After many months, it became clear that the government and Telstra

were unable to reach agreement. In August 2006, Telstra announced that

it would not continue with its proposed network upgrade as it had not

reached agreement on pricing issues. In a share off er prospectus released

in 2006, Telstra commented:

Telstra seeks a competitive rate of return when it invests its capital. If Telstra cannot be confi dent that ACCC regulation of prices for competitor access to a new network will allow a competitive rate of return, Telstra will not invest in the network. This year, Telstra planned to start building a $3 billion FTTN network. However, Telstra disagreed with the ACCC on the price its competi-tors should pay for access to the network and, as a result, Telstra decided not to build the network.

Meanwhile, a consortium comprised of Telstra’s competitors (known

as ‘G9’) announced a rival proposal to build its own FTTN network. The

G9 proposal required the government to enact legislation that prohib-

ited Telstra overbuild and required the forced cutover of existing Telstra

copper infrastructure to the new network at cost- based rates. By requiring

a prohibition on competitive overbuild, the G9 proposal eff ectively sought

to create a monopoly franchise in its favor. The G9 proposal was also con-

ditional on agreement of an undertaking with the ACCC on access prices

to reduce regulatory risk.

During 2007, the FTTN network became a signifi cant issue in Australia’s

federal election. The Minister for Communications announced that a

tender process would be established. Few parties subsequently expressed

serious interest, leaving G9 and Telstra as the only potential bidders. Some

notable investors cited excessive regulation as a deterrent to investment.

The Liberal government lost Australia’s December 2007 federal election

and a new Labor government came to power. The new minister terminated

the previous FTTN tender procedure and substituted a new procedure.

The new tender sought proposals to roll out an Australian national broad-

band network (as a NGN) that would cover 98 percent of the population

with a minimum bandwidth of 12Mbit/s into the home. The government

indicated that it would make a funding contribution of up to AU$4.7

billion to the total estimated cost of AU$15 billion.

In November 2008, six parties submitted proposals to government

for the NGN. Telstra refused to submit its full proposal on the basis

that the government had not provided suffi cient regulatory certainty.

Controversially, Telstra’s proposal was disqualifi ed in December 2008

due to non- compliance with tender requirements.The government is now

building its own fi ber to the premises network at very substantial cost,

eff ectively replicating Telstra’s existing access network.


3.4.4 Insights from the Australian Case Study

In Australia, regulation has persisted, even increased in some cases, as

competition has increased. Such regulation has deterred investment by

the incumbent. Such regulation has also deterred long- term infrastruc-

ture investment by market entrants by enabling them to obtain access to

existing infrastructure at a materially lower cost than if they rolled out

infrastructure themselves. While this has promoted short- term market

entry, it has created a market dependence on continued access regula-

tion at the expense of the development of long- term infrastructure- based

competition.

Australia recognized the problems that Part XIC created for investment

at an early stage and sought to create an access holiday regime. However,

investment is sensitive to uncertainty and time delays and the new regime

suff ered both problems. As a consequence, Telstra sought to bypass Part

XIC by negotiating directly with the government when upgrading its infra-

structure. However, the continued emphasis by the ACCC on TSLRIC

has led to an impasse and no network rollout.

Ultimately, the issue became highly politicized. Infrastructure rollout

was proposed via a public–private partnership (PPP) supported by partial

government funding and amendments to the current regulatory settings.

However, following a further impasse on these issues, Telstra was dis-

qualifi ed from the NGN tender procedure. As Telstra was widely viewed

as the only entity capable of rolling out the NGN to the government’s

specifi cations, the likelihood of an NGN rollout in Australia remains

uncertain.

Given this outcome, the existing Australian regulatory regime has

undeniably failed to achieve one of its stated objectives of ‘achieving eco-

nomically effi cient investment in infrastructure’ (Trade Practices Act 1974

(Cth), s 152AB(2)(e)).

3.5 TECHNIQUES TO PROMOTE INFRASTRUCTURE INVESTMENT

Appropriate incentives can be created to encourage large- scale infrastruc-

ture investment by reducing the level of regulatory risk. Of the various

mechanisms adopted in other industries, the most common are:

the sharing of regulatory risk between the public and private sector ●

via binding governmental commitments (enforced by compensa-

tion), as frequently occurs in public–private partnerships; and


access holidays, typically providing certainty that a network owner ●

will earn a positive NPV suffi cient to justify the infrastructure roll-

out.

Both of these techniques are considered in further detail below. Project

fi nancing structures also utilize techniques such as third party insurance

as a means of mitigating regulatory risk for fi nanciers. Such insurance is

usually known as ‘political risk insurance’ and is normally directed at the

most extreme forms of regulatory risk, such as expropriation of assets and

currency inconvertibility.

3.5.1 The Public–Private Partnership Model

Large- scale infrastructure projects involve signifi cant risks. A wide range

of stakeholders are involved in such projects, whether by contributing

debt or equity, or by contributing goods or services. An underlying prin-

ciple guiding the negotiation of contracts between these parties is that the

project risks should be allocated to those parties best placed to control or

manage them. In this manner, the level of risk should be reduced and the

NPV of the project should increase.

A recent legal innovation in many countries has been the formal devel-

opment of integrated PPPs to reduce the fi nancial burden on the public

sector while increasing the effi ciency of service delivery. A PPP is best

described as a set of contractual arrangements between public and private

sector entities to deliver cost- eff ective and high- quality services to the

public over an extended time period. Each party contributes its skills and

expertise and shares in the risks and rewards.

Importantly, PPPs seek to implement a key principle of risk redistri-

bution. Participating entities are allocated those risks that they are best

placed to control or manage. Under this principle, a party will bear a risk

where it is within that party’s control and/or which it can most effi ciently

manage relative to other parties (including by spreading the risk to third

parties). In eff ect, project risks are allocated so that each party bears those

risks that it can best control and/or manage and at the least overall cost

to the project.

Relevantly, the government normally bears a substantial proportion of

the regulatory risk associated with large PPP projects under the terms of

the relevant ‘concession agreement’. This is primarily because:

Regulatory risks arise as a result of governmental action, hence ●

the government will normally have the political and legal power to

prevent adverse regulatory events occurring.


If adverse regulatory events do occur, the government can most ●

effi ciently bear such risks by pooling and spreading the cost of com-

pensation across taxpayers. Such spreading of risk to the public is

appropriate, for example, if the government has determined that

regulation is required in the public interest.

The key insight from PPPs is that it is possible to reduce regulatory risk

via a binding contract between the public and private sector. The private

sector will agree to roll out infrastructure in the public interest with agreed

pricing that provides a positive NPV. In consideration, the government

will forbear imposing certain forms of regulation on that network (for

example, price controls) and will bind itself to pay compensation if this

commitment is breached.

A key characteristic of NGN telecommunications projects is their high

degree of exposure to regulatory risk. As evidenced by the Australian

experience, such risks can be suffi ciently great as to render an NGN project

uneconomic. By reducing the regulatory risk via a PPP arrangement, the

NPV of the project will increase and it will be more likely to proceed.

In essence, an NGN project will be more likely to proceed if the

private sector can obtain a meaningful binding commitment (enforced

via compensation) from government that it will not be subjected to value-

destructive regulation that will result in a negative NPV.

3.5.2 Access Holidays

A specifi c form of binding commitment given by the government to the

private sector to promote investment is an ‘access holiday’. An access

holiday is a defi ned period of time within which new infrastructure invest-

ment is exempted from some or all access regulation. Typically, the access

holiday mechanism is enacted into law as a statutory exemption to an

existing legislated access regime.

The benefi ts of an access holiday were summarized in the 2003 article by

Professors Gans and King in the following terms:

If regulators are expected to set low access prices ex post but, at the same time, there is a commitment that any new essential facility will not be subject to infrastructure access for a signifi cant period of time, then this will raise investor incentives. It is this ex ante commitment to delay access ex post that is the basis of an access holiday. In this sense, access holidays play a role similar to a patent in innovative activity. Patents encourage innovations by conferring on the inventor tem-porary monopoly profi ts. Similarly, access holidays encourage infrastructure investment by allowing investors to temporarily exploit any market power


associated with their facility. Both patents and access holidays are second- best solutions in that they impose a temporary monopoly cost. Both an optimal patent and access holiday needs to be designed to trade- off this temporary loss with increased incentives to invest. (Gans and King, 2003, p. 163)

The Gans and King analogy of a patent can be extended:

A policy rationale for granting exclusive rights to patent holders is ●

that the patented innovation may have spillover positive externality

benefi ts for wider society. In this manner, the patent holder is, in

eff ect, being allocated (via a monopoly right) a proportion of the

wider utility benefi t that accrues to wider society.

The same can be said in relation to access holidays for new telecom- ●

munications networks. Such networks have similar positive spillover

externalities for wider society, explaining why governments are so

keen on promoting them. It is therefore appropriate that network

owners receive an additional reward in order to stimulate such

investment.

As indicated by the patent analogy, the central policy concern in confer-

ring an access holiday is the ability of the infrastructure owner to utilize

any pre- existing substantial market power that it may possess. Specifi cally,

how should policy- makers balance the promotion of infrastructure invest-

ment against the prevention of misuses of market power? This issue of

an optimal access holiday has been debated in Australia in the context of

investments in natural gas pipeline infrastructure.

In August 2004, the Australian Productivity Commission released a fi nal

report on its review of Australia’s national third party access regime for

gas pipelines. The Productivity Commission identifi ed that access pricing

for gas pipelines was impeding investment and innovation. It reasoned

that the existing regime was excessively focused on cost- based price regula-

tion and this caused a truncation of returns that deterred investment. For

gas pipeline infrastructure that did not confer substantial market power,

the Productivity Commission recommended the introduction of a 15- year

access holiday regime.

However, the Productivity Commission did not recommend an access

holiday recommendation for gas pipeline infrastructure that did confer sub-

stantial market power. Following signifi cant resulting concern expressed

by pipeline owners, the Australian Ministerial Council on Energy pro-

posed a compromise: a partial access holiday would apply to pipelines

with substantial market power in which access prices would be exempted

from regulation but all non- price terms would remain regulated.


Following signifi cant concern expressed by pipeline users, a negotiated

political compromise was ultimately achieved that was supported by both

pipeline owners and pipeline users:

All new pipelines without market power were permitted to apply for ●

a 15- year holiday from all access regulation.

New international pipelines with market power were permitted to ●

apply for a 15- year holiday from price regulation, but would remain

subject to non- price regulation.

New domestic pipelines with market power would have no right to ●

an access holiday, but would instead be subject to price regulation

triggered by a higher regulatory threshold.

This compromise was subsequently enacted into Australian law.

Interestingly, this compromise may still deter investment in those domes-

tic pipelines that would clearly meet the higher regulatory threshold. The

Gans and King solution would suggest that such pipelines should be guar-

anteed some kind of positive NPV return suffi cient to create an incentive

for investment and reward the investor.

3.5.3 Policy Recommendations to Promote Greenfi eld Investment

Several potential policy recommendations for NGN infrastructure can

therefore be drawn from the analysis set out in this chapter. Firstly it is

potentially benefi cial for government to grant an access holiday to encour-

age investment in new telecommunications infrastructure that will not

confer substantial market power on the network owner. This may be the

case, for example, if the NGN would not be the only means of delivering

broadband Internet access into the home due to intermodal competition.

Secondly, if new telecommunications infrastructure would confer sub-

stantial market power on the network owner, it is still potentially benefi cial

to provide a more limited access holiday arrangement. Under this arrange-

ment, the network owner would not be subjected to TELRIC pricing (and

its disincentives) but would instead earn a regulated return that generated

a value- accretive positive NPV suffi cient to incentivize the investment. In

this manner, any regulated access pricing should be increased well above

the TELRIC levels that have traditionally applied.

Thirdly, particular jurisdictions may be susceptible to changes in policy

so that ‘access holiday’ rights could be repealed by a subsequent govern-

ment. In such circumstances, it may be appropriate for network owners

also to consider entering into an agreement with government in which

compensation is paid by the government if an access holiday is prematurely


removed. However, any such entitlement to compensation would likely be

conditional on the network achieving certain quality standards, possibly

under a PPP arrangement.

3.6 CONCLUSIONS AND RECOMMENDATIONS

A key focus of telecommunications regulatory policy at present concerns

the impact of regulation on the long- term incentives for new infrastruc-

ture investment. This chapter has analyzed the theory and practice behind

the regulation versus investment debate and made a series of key policy

recommendations.

In summary, legacy customer access network infrastructure has its

technological limitations and therefore substantial investment in NGN

infrastructure will be required. However, investment in such infrastructure

is inherently risky because of sunk costs and long asset lifetimes. Under

modern fi nance theory, such investments will only proceed if the network

owner can make a return on its investment that exceeds its cost of capital

over the project life. The rate of return will depend on the risk profi le of a

project. In the case of an NGN project, the most critical project risks are

often regulatory risks.

Against this context, international best practice in telecommunications

regulation has historically favored a high degree of network unbundling

and the application of cost- based access regulation. While such regula-

tion has promoted market entry and services- based competition, it has

impeded infrastructure investment and the development of long- term

facilities- based competition. The optimal regulatory approach should be

to wind back sectoral regulation as competition develops so as to preserve

long- term network investment incentives. In the absence of such wind-

back, low access prices may result in telecommunications infrastructure

not being built.

The genesis of the network unbundling concerns the price at which

access is provided to unbundled network elements. Incentive- based

TELRIC pricing was intended to balance allocative and dynamic effi -

ciency considerations, but it has clearly had a disproportionate adverse

eff ect on dynamic effi ciency. Generally, such pricing has capped the upside

of telecommunications investment while exacerbating the downside risks.

Furthermore, competitors can free- ride on the investor’s infrastructure at

a lower cost than rolling out their own infrastructure. Serious policy ques-

tions therefore arise regarding the appropriateness of applying TELRIC

pricing to greenfi eld investments and network upgrades.

A case study of the proposed rollout of NGN infrastructure in


Australia provides some practical insights into these issues and suggests

some policy solutions. Generally, regulation has persisted in Australia,

even increased, as competition has developed. Such regulation has had

an adverse eff ect on investment. Australia sought to correct the adverse

incentive eff ects of access regulation at an early stage by introducing an

access holiday regime. However, for various reasons, this attempt was

unsuccessful. Given the extent to which the access regime has deterred the

rollout of the Australian NGN, the Australian government intervened by

off ering partial government funding; in eff ect, proposing a PPP arrange-

ment. Ultimately, the government has been required to make the invest-

ment itself.

As evidenced by the attempts in Australia, the general policy solution

to the investment problem is to generate suffi cient and certain returns for

investors that they are guaranteed a value- creative positive NPV over the

long lifetime of their investment. However, if the infrastructure investment

also confers substantial market power on the network owner, it will also

be necessary to implement a policy solution that prevents the misuse of

that market power over the investment lifetime.

This chapter concludes that the appropriate pricing methodology to be

applied in that scenario is not an incentive- based TELRIC price. Rather,

the price methodology should provide a return to the investor in the nature

of a patent: the investor should earn suffi ciently above its WACC that the

investment becomes highly desirable. Such higher returns to investors are

justifi able given that telecommunications networks have positive spillover

externalities for wider society.

In summary, nations need to adopt a more appropriate regulatory

balance that appropriately rewards investors for investment in new tel-

ecommunications infrastructure. Only if a better balance is achieved will

we achieve a vision for the year 2015 of ubiquitous broadband Internet

access.

NOTE

1. The views in this chapter refl ect the personal views of the author and not necessarily the views of any of his current or previous clients.

REFERENCES

Australian Productivity Commission (2001). Telecommunications com-petition regulation: Productivity Commission inquiry report. Retrieved 10


March 2009 from http://www.pc.gov.au/inquiry/telecommunications/docs/fi nalreportcfm?abstract_id=996065.

Crandall, R. and G. Sidak (2007). Is mandatory unbundling the key to increasing broadband penetration in Mexico? A survey of international evidence. Social Science Research Network, Retrieved 10 March 2009 from http://papers.ssrn.com/sol3/papers.

FCC (Federal Communications Commission) (2003a). In the Matter of Review of the Section 251 Unbundling Obligations of Incumbent Local Exchange Carriers (CC Docket No. 01- 338). Report and order on demand and further notice of proposed rulemaking, released 21 August.

FCC (Federal Communications Commission) (2003b). In the Matter of the Commission’s Rules Regarding the Pricing of Unbundled Network Elements and the Resale of Service by Incumbent Local Exchange Carriers (WC Docket No. 03- 173). Notice of proposed rulemaking, released 15 September (Triennial Review Order).

Gans, J. and S. King (2003). Access holidays for network infrastructure invest-ment. Agenda, 10(2), 163.

Gayle, P. and D. Weisman (2007). Effi ciency trade- off s in the design of competi-tion policy for the telecommunications industry. Review of Network Economics, 321(6), 321–41.

Mandy, D. and W. Sharkey (2003). Dynamic pricing and investment from static proxy models. Review of Network Economics, 4(2), 403–37.

Sidak, G. (2006). A consumer- welfare approach to network neutrality regulation of the Internet. Journal of Competition Law and Economics, 4(55), 55–99.

Telstra Corporation Ltd. (2006). Telstra 3 Share Off er: share in the future. Prospectus, 9 October. Retrieved 10 March 2009 from http://www.telstra.com.au/abouttelstra/investor/prospectus.cfm.

83

4. Behavioral economics and telecommunications policy

Patrick Xavier and Dimitri Ypsilanti

4.1 INTRODUCTION

Regulation in the telecommunications sector has focused mainly on

the supply side of the market including, for example, market entry and

licensing, access to and use of networks, interconnection, control over

retail and/or wholesale pricing. This emphasis on the supply side has

been appropriate because the task was to install eff ectively competing

alternative suppliers in former monopoly telecommunication markets.

As competition has developed in telecommunication markets and users

have a wider choice of service providers, there has been increased atten-

tion by some regulators on the demand side. For instance, a demand- side

measure introduced in many countries is the requirement for ‘number

portability’ aimed at facilitating consumer ‘switching’ in the fi xed line and

mobile markets.

Such attention to the consumer demand side is timely, because informed

consumers prepared to choose between competing suppliers are necessary

to stimulate fi rms to innovate, improve quality and compete in terms of

prices. Indeed, in making well- informed choices between suppliers, con-

sumers not only benefi t from competition, but they exert the sustained

pressure for providers to compete for their customer. Conversely, where

consumers have too little information, poor- quality information, or

misinformation, they may end up misled and confused by the choices on

off er, may pay too much or may buy the wrong service. This may, in turn,

inhibit and dampen the competitive process. Moreover, if suppliers can

exploit consumers this could at some point lead to ‘reputational damage’

and disillusionment with the competitive process as a whole. Hence con-

sumer protection is critically important from an economic as well as social

rationale.


4.1.1 Telecommunication Policy and Regulation in the Interests of the

Consumer

Policy and regulation in the telecommunications sector has long been said

to be concerned with the consumer interest (including consumer protec-

tion and empowerment). Legislation, policy and regulatory statements

repeatedly claim that the welfare of consumers is a, if not the, primary

objective of policy and regulation. But, despite this rhetoric, there has

been relatively little analysis of the extent to which consumers have in fact

benefi ted and become more empowered. Competition among suppliers to

attract and retain consumers was supposed to result in suppliers that serve

consumers well prospering relative to those that do not.

Now with market liberalization and alternative telecommunications

suppliers becoming fi rmly installed in many countries, questions are

emerging about whether consumers have in fact become empowered

and ‘switched’ to alternative suppliers as much as had been expected. If

not, why not? Should policy- makers and regulators intervene? And if so,

how? In seeking to answer these questions, important information can be

gleaned from analysis of the demand side, including new insights from

behavioral economics into actual (as distinct from normative) consumer

behavior.

4.1.2 Demand- side analysis, including behavioral economics

Conventional economics starts with a number of assumptions about con-

sumer behavior, including that:

Consumers approach markets with a set and stable set of ●

preferences.

Consumers are concerned only with their own welfare. ●

In aggregate at least, consumers can rationally use available infor- ●

mation to make optimal decisions.

Demand- side analysis, including so- called behavioral economics, goes

beyond these assumptions. It argues that in some situations consum-

ers consistently depart from behavior predicted by these assumptions.

In other words, consumer choice is biased away from that which

would occur if behavior conformed to the assumption of ‘rationality’.

Behavioral economics accords particular reference to the observed, as

opposed to theorized, behavior of consumers (Pesendorfer, 2006). A key

issue for a demand- side analysis is examining ‘what is actually going on

in terms of consumer outcomes’ (Sylvan, 2006b). This involves examining

Behavioral economics and telecommunications policy 85

the product features as well as consumer choice issues and how the

product or service is being off ered or presented. An important insight of

behavioral economics is that it is not only the product that matters, but

the context as well.

The conventional economic approach recognizes that information

asymmetry and information failure may lead to suboptimal consumer

outcomes. It also recognizes that consumers face a ‘bounded rationality’

(for example, in the form of various costs of acquiring and processing

information) and, as a result, rely on ‘heuristics’ (such as reliance on a

fi rm’s reputation and other price and quality signals) in decision- making.

Consumer policy has therefore focused on removing those market failures

that prevent consumers who, even though they behave rationally, are

nevertheless prevented from making optimal choices that maximize their

welfare.

A substantial amount of regulation has been directed at these problems,

including implementing misleading- conduct laws (for example, prohibi-

tion on fi ne print disclaimers, requirement for plain language contracts);

disclosure requirements; product regulation; and regulation aimed at

allowing consumers to switch conveniently between suppliers – such as

the development of interoperability standards and number portability

(as discussed later) in the telecommunications industry. Other potentially

useful tools include the provision of price and quality comparison data

for consumers, and the use of calculators (technical tools usually on

websites) allowing consumers to enter data to enable them to make price

comparisons.

Behavioral economics challenges some of the assumptions of conven-

tional economics that consumers make their choices coherently and ration-

ally given their preferences and the constraints upon them. Behavioral

economics involves the study of actual consumer behavior (by contrast

with how consumers should behave) (OECD, 2008). It argues that con-

sumers often fail to act in their own best interests due to behavioral traits

such as failure to process information objectively or misevaluations about

the costs and benefi ts of prospective decisions. For example, research

conducted on switching activity by low- income households in UK elec-

tricity markets found that on the whole these consumers did not switch

provider in a way that could be explained by any rational set of criteria.

Only 7 percent of consumers chose the cheapest option and, indeed, 32

percent changed to a supplier that was more expensive (Waddams- Price

and Wilson, 2005).

Among the biases identifi ed by behavioral economics, the following

may be particularly relevant to the telecommunications market and may

help to explain how, even where there is adequate information, consumers


may be making seemingly irrational decisions in choosing an operator or

service package (OECD, 2006).

Choice overload

Consumers having too many products or features to compare and may

experience increased anxiety about the possibility of making a bad choice.

This can lead to random choice, or failure to make any choice, resulting in

missed opportunities for buyers and sellers. A type of ‘analysis paralysis’

can take hold when information and choice become very complex.

Endowment

Consumers may be reluctant to give up what they have, even though they

would not buy such goods or services if they did not already have them

(for example, consumers may stay with the incumbent fi xed line provider

because of misplaced loyalty, a failure to acknowledge poor choices in the

past, or an irrational consideration of sunk costs).

Defaults

The ordering of options, particularly in markets where a choice must

be made, infl uences choice. Default bias means the decision to opt- in or

opt- out (for example, of extending a mobile phone contract) is not the

same decision for all people. Also, consumers may tend to take a path of

least resistance, particularly if they feel that there is a ‘normal’ option (for

example, people may buy ‘standard’ bundles off ered by telecommunica-

tions suppliers, even if they do not want the whole telecommunications

bundle).

Hyperbolic discounting

Consumers tend to be short- sighted when making decisions with immediate

costs or benefi ts to be weighed against future costs or benefi ts (for example,

consumers may enter long- term telecommunications contracts because

they place more value on the immediate benefi ts of the off er, such as a free

or heavily subsidized handset or a reduced fi rst month rate, or free local

calls, rather than on the long- term costs of a contract such as high prices

for calls exceeding a usage ‘bucket’, the inability to switch to lower- priced

alternatives, and the inability to take advantage of latest technology).

Time- variant preferences

This usually manifests as hyperbolic discounting. Generally, consumers’

discount rate increases as the outstanding time period becomes shorter

and this leads to decisions which discount costs (which are incurred over

the longer term) more than benefi ts (which are realized in the short term).


This can lead consumers to sign long- term contracts with mobile phone

companies to obtain a free handset, even though this may be more expen-

sive in the long term than buying a handset without being locked in to a

long- term contract.

Framing biases

Consumer choice is infl uenced by the ‘frame’ in which information is

presented. Presentation of the same information in a diff erent frame can

lead to a diff erent decision. For instance, cash- back off ers can be much

more attractive to customers than a similar or even greater discount. For

most consumers, ‘only 3 percent fat’ is likely to be less appealing than ‘97

percent fat free’.

Loss aversion

The preference for avoiding loss is widely considered to be greater than the

preference for gain.

Heuristics

Consumers often take short cuts (for example, by following rules of

thumb) when the decision environment is too complex relative to their

mental and computational capabilities. These rules of thumb are called

‘heuristics’ and are often accurate enough to be useful, but may sometimes

lead to suboptimal decisions.

4.2 INFORMATION IMPERFECTIONS

Where consumers have too little information or too much information of

inconsistent quality, various types of problem can arise, including:

consumers failing to participate in the market at all, because they ●

have limited awareness of the products and services on off er, or con-

versely because they are confused by an excess of available informa-

tion (sometimes referred to as ‘information overload’);

consumers paying too much; ●

consumers buying the wrong product or service; ●

consumer disappointment with the product or service, because it ●

turns out to lack the expected level of quality (Ofcom, 2006a).

Constraints on the ability of consumers to process information can

also lead to non- optimal, welfare- reducing decisions even when the infor-

mation available to them is plentiful and non- deceptive. For example,


comparison of prices for telecommunications services off ered by diff erent

suppliers is complicated by the wide range of possible consumer usage pat-

terns, detailed variations in price levels and price structures, and the large

number of possible discount and bundled schemes available. Moreover,

operators and service providers regularly adjust their pricing strategies

either through changes in components of their tariff structure, or the intro-

duction or withdrawal of various discount schemes and service packages

(Miravete, 2007).

Technological advances and market pressures have made telecommuni-

cations and information and communication technology (ICT) products

and systems increasingly complex and feature- rich. But this has also

increased the complexity faced by consumers in determining appropriate

products that satisfy consumer needs and in evaluating alternative product

off erings by competing providers.

4.2.1 Ability and Willingness to Switch

Barriers to switching service providers can be present due to high switch-

ing costs. Switching costs can be defi ned as the real or perceived costs

that are incurred when changing supplier but which are not incurred by

remaining with the current supplier.

The ability and willingness of consumers to switch is critically impor-

tant. If switching is discouraged or impeded this could impact not only on

the demand side but also potentially raise supply- side barriers (Barrow,

2007). This is because new entrants could be deterred from entering the

market in the belief that it will be diffi cult to persuade consumers to switch

from their existing provider. This could diminish contestability and the

eff ectiveness of competition and limit the benefi ts that consumers would

otherwise derive from it.

In the UK, research conducted by Ofcom (2006b, 2008), the UK

National Consumers Council (2006) and others suggests that in the tel-

ecommunications sector there are, in fact, a range of important deterrents

to switching, including:

Lengthy and cumbersome switching procedures that make it incon- ●

venient for consumers to switch and can outweigh any potential

benefi ts.

Early exit charges, imposed by an existing provider, that reduce the ●

benefi ts of switching.

Confusing products and non- transparent pricing that makes it dif- ●

fi cult or time- consuming to compare deals (as in the case of mobile

telephony and the Internet).


Technical incompatibility of equipment that can make it uneco- ●

nomical for consumers to switch (for example, if they cannot use a

blocked mobile phone with their new provider).

Long- term deals that lock consumers into lengthy relationships with ●

their providers (as may occur with mobile telephony and Internet

contracts) and increase the risk of them being overcharged.

Consumers will not switch to a competing supplier unless the price dif-

ference exceeds the switching costs. Where switching costs are high, it is

possible for a provider to set very low prices, even prices below cost to

attract new consumers, but then subsequently charge these consumers

prices well above costs once they are ‘locked in’. This ‘bargain then rip- off ’

pricing pattern is a characteristic of many markets with high switching

costs (NERA, 2003).

4.3 EVIDENCE OF ACTUAL CONSUMER BEHAVIOR IN TELECOMMUNICATIONS MARKETS

As noted earlier, behavioral economics involves the study of actual

consumer behavior as distinct from how consumers should behave.

Accordingly, this section examines the available evidence of actual con-

sumer behavior in the telecommunications services sector and infl uences

on such behavior.

4.3.1 Consumer Behavior in the Fixed Line Market

United Kingdom

Ofcom’s research (2008) found that in the United Kingdom, one in three

(34 percent) of the fi xed line consumers surveyed had changed the supplier

providing their home fi xed- line service between 2003 and 2006. However,

two- thirds (66 percent) had not switched in the 2003–06 period and the

majority (52 percent) had not even considered doing so. Regardless of

whether or not they had switched supplier in the last four years, 34 percent

had made some change to their existing service with their current supplier

in 2003–06. A similar proportion claimed to keep an eye on the market:

around a third (36 percent) agreed that they were always on the look- out

for a better deal and a similar proportion (31 percent) agreed that they

made a conscious eff ort to keep up to date with what other providers were

off ering. In this context, it is notable that according to Ofcom’s research,

86 percent of fi xed line consumers in the UK were satisfi ed with their


overall experience with their current supplier: almost half (46 percent)

were very satisfi ed and a similar proportion (40 percent) were fairly satis-

fi ed. Clearly, if consumers are satisfi ed, there may be no or less reason for

them to switch.

The fi gures in Table 4.1, showing that well over 70 percent of fi xed line

consumers in the UK did not switch suppliers during 2007 and 2008, are

broadly consistent with the data for earlier years.

Consumer responses to Ofcom’s questionnaire surveys indicated that

the greatest deterrent to shopping for an alternative fi xed line supplier is

the possibility of getting locked into a contract with a new supplier: two-

thirds of consumers (67 percent) agreed that they would be put off by this.

The second- greatest barrier to switching was reluctance to leave a known

and trusted supplier for one that is unfamiliar – stated by 65 percent of

consumers. In the fi xed line market, many consumers have been with their

supplier for many years. For some consumers, their relationship with their

fi xed line supplier was perceived as being more important than whether

they could get a better deal elsewhere. Because of the ‘endowment factor’,

a factor infl uencing consumer behavior underlined by behavioral econom-

ics, these consumers would only switch if they had experienced a serious

betrayal of trust that incites a revenge value to switching. As many as one

in two (53 percent) fi xed line consumers agreed that they had a strong

sense of loyalty to their existing supplier (Ofcom, 2006b). No signifi cant

diff erences were detected between demographic groups. In other words,

diff erences are due to attitudes and behavior rather than demographics.

In summary, Ofcom concludes that in the fi xed line market, in the face

of an undiff erentiated market (perceived or actual), some consumers are

adopting fallback, risk- averse strategies and will stay with what they

know and trust, even though it might not be the best rational option.

Table 4.1 Switching in the UK fi xed line communications market during

2007 and 2008

2007 (%) 2008 (%)

Switched supplier 9 12

Actively looking 3 4

Started looking, not switched 4 5

Considered without looking 7 6

None 77 73

100 100

Source: Ofcom (2008).


This appears to support the arguments of behavioral economics. Ofcom’s

fi ndings suggest that the greatest chance of furthering participation in the

fi xed line market are eff orts to help consumers to overcome their inertia,

by allaying fears regarding the potential risks associated with switching

services, or by educating them regarding the tangible benefi ts of any new

service over and above their existing arrangement (Ofcom, 2006b).

Portugal

The level of switching in Portugal’s fi xed line market has also been

relatively low. Table 4.2 indicates that about two- thirds (64 percent) of

consumers interviewed as part of an Anacom (2006) (the telecommunica-

tions regulator) study were satisfi ed with the prices charged by their fi xed

network provider. This degree of satisfaction with a key aspect of service

would clearly reduce the disposition to switch.

4.3.2 Consumer Behavior in Mobile Telecommunications Markets

United Kingdom

Ofcom found that more than a third (36 percent) of mobile telecommu-

nications consumers in the UK had changed their mobile phone network

supplier during 2003–06. The fi gure rises to 52 percent for those who are

on an annual contract and falls to 29 percent for those who are on pre- pay

packages. However, two- thirds (66 percent) had not switched during this

time period and the majority (53 percent) had not even considered doing

so. Even among the 7 percent who had considered switching, not all had

actively started looking for an alternative.

The fi gures in Table 4.3 show that well over 70 percent of mobile con-

sumers in the UK did not switch suppliers during 2007 and 2008, and this

is broadly consistent with the data for earlier years.

Ofcom (2006b) pointed out that although the majority of consumers

in the mobile phone market have not considered changing their supplier,

Table 4.2 Degree of satisfaction towards fi xed line prices in Portugal

Degree of satisfaction %

Very satisfi ed 5.8

Satisfi ed 58.2

Not satisfi ed 29.5

Not satisfi ed at all 4.1

Do not know/Did not respond 2.5

Source: ANACOM (2006).


they are nevertheless participating in other ways: 32 percent had made

some change to their existing service with their current supplier during

this time period. While consumers were far more likely to change the tariff

or package they were on (31 percent) than ask their supplier to match a

better deal they had seen elsewhere (8 percent), mobile phone users seemed

aware of their potential to negotiate: 62 percent of those on contracts had

changed their existing tariff or package and 16 percent had attempted to

renegotiate their package or deal. Around a third of all consumers claimed

to keep an eye on the mobile phone market; 32 percent agreed that they

were always searching for a better deal and 28 percent agreed that they

made a conscious eff ort to keep up to date with what other providers were

off ering. In terms of reasons given for not switching, more than half the

telecommunications consumers surveyed (54 percent) said that they were

very satisfi ed with their overall experience with their current supplier and

another third (36 percent) said that they were fairly satisfi ed. Post- pay

and pre- pay consumers were equally satisfi ed with their current network

supplier.

The research also identifi ed a range of procedural and psychological

barriers perceived by mobile telecommunications consumers as Table 4.4

indicates. The greatest deterrent to shopping for an alternative mobile

phone network supplier was the possibility of getting locked into a con-

tract with a new company. Around two- thirds of consumers (68 percent)

agreed that they would be put off by this.

Refl ecting a similar sentiment, the second- greatest barrier to shopping

was reluctance to leave a known and trusted supplier for one that was

unfamiliar, an issue for 64 percent of consumers. This refers to the extent

to which consumers value their relationships with their mobile phone

network supplier. More than one in two (55 percent) agreed that they had

Table 4.3 Switching in the UK mobile communications market during

2007 and 2008

2007 (%) 2008 (%)





None 74 74

100 100



Table 4.4 Perceived barriers to shopping and/or switching mobile phone

suppliers in the UK

Response to question (with prompted reasons): ‘I did not switch my mobile phone

supplier because . . .’

Reason Agree

(%)

Strongly

agree (%)

Total

agree

(%)

Post- pay

(%)

Pre- pay

(%)

Don’t want to get locked

into contract with new

provider

47 21 68 47 52

Reluctant to leave provider I

trust for one I don’t know

47 17 64 52 59

Strong sense of loyalty

towards current provider

36 19 55 No diff erence

Shopping for new provider

too much of a hassle/chore

41 12 53 41 49

Only short term gain as

providers follow each

other


Diffi cult to make

comparisons b/w

providers

38 9 47 35 53

Don’t have time to research

options

38 9 47 34 54

Don’t want to lose current

deal/package


Big risk that something will

go wrong in transition


Don’t know enough to make

right choice

28 9 37 27 42

Bound to feel stupid/out- of-

date when I talk to sales

staff

21 12 33 24 37

No diff erence in cost of

supplier

28 4 32 27 33

No diff erence in quality of

supplier

26 3 29 23 32

Don’t know where to fi nd

trusted info about options


Note: Base: all mobile (500). Percentages represent share of total sample.

Source: Ofcom (2006b).


a strong sense of loyalty to their existing network supplier. This was par-

ticularly the case for inactive consumers of whom 39 percent were likely to

be relational people (with a preference for using trusted brands known or

recommended to them) and only 14 percent were likely to be transactional

people (willing to consider unfamiliar brands if they off er a good deal).

These manifestations of the endowment factor and/or irrational fear or

risk- averseness relating to a change to a new provider are consistent with

the arguments of behavioral economics.

More regulatory attention might be directed to this ‘barrier to switch-

ing’. For instance, the cost of breaking a contract might be lowered where

a consumer can demonstrate that conditions of the contract were not

being delivered by the supplier.

The consumer traits of lack of confi dence, heuristics and susceptibil-

ity to information overload emphasized by behavioral economics also

appeared to play a role in decision- making. Among so- called inactive con-

sumers, 48 percent did not feel they knew enough to make the right choice

and 42 percent expressed concern about appearing stupid in front of sales

staff , while 44 percent were willing to accept a solution that they felt was

good enough rather than investigate all options to fi nd the best one.

In this market, consumers on contracts also appear to be bound to

their suppliers as a result of having negotiated or been given special deals:

two- fi fths (42 percent) of consumers expressed concern about losing the

package or deal they were on. Early- exit penalties could also apply here.

The process of shopping was itself likely to discourage around half of

the consumers surveyed: 53 percent agreed that shopping for a new sup-

plier was too much of a hassle; 47 percent agreed that it was diffi cult to

make comparisons between suppliers; and 47 percent agreed that they

did not have enough time to research the options. Moreover, many (52

percent) perceived that the gain would be short term because all the sup-

pliers follow each other. As Table 4.4 indicates, this is particularly the

case among pre- pay consumers, who perceived higher barriers to switch-

ing than those on contracts: 82 percent of pre- pay users agreed they did

not want to be locked into contracts with a new provider. They were also

signifi cantly more likely to regard shopping for a new network supplier as

an onerous process, approaching it with lower levels of interest and confi -

dence than contract users.

In summary, Ofcom concluded that the evidence suggests that in the

UK mobile telecommunications market, in the face of complexity and

lack of market diff erentiation (perceived or actual), some consumers will

stay with what they know, even though it might not be the ‘best’ option.

Ofcom concluded that the greatest chance of furthering consumer partici-

pation in the mobile phone market is through eff orts made to infl uence


these key drivers, for example by educating them in regard to the tangi-

ble benefi ts of any new service over and above their existing set- up, by

making the switching process easier or by helping consumers to overcome

inertia.

Portugal

In Portugal, Table 4.5 indicates that of those mobile telecommunica-

tions consumers responding to a survey conducted for Anacom about 19

percent have switched. Of these, about one- third switched because most

of their contacts were clients of the new operator. Presumably there was

some advantage with being on the same network such as network cover-

age and/or discounted or free calls to consumers on the same network.

Another one- third of those who switched did so because they were

unhappy with the prices of their old operator; while 13.7 percent of those

who switched did so because they were unhappy with the quality of service

provided by their old operator, and 8.6 percent because of an off er from

the new operator.

Of the 80.7 percent who had never changed operator, 66 percent were

satisfi ed with their current operator and 31.2 percent because most of their

contacts are clients of their current operator. Only about 10 percent of

Table 4.5 Switching mobile telecommunications operator in Portugal

Reasons to change operator %

Have already changed operator 19.3

Most of my contacts are clients of new operator 34.6

Unhappy with prices 33.3

Unhappy with quality of service 13.7

Off er from new operator 8.6

Other 20.6

Never changed operator 80.7

Satisfi ed with current operator 66.2

Most of my contacts are clients of current operator 31.2

Off ers the best prices 6.6

Switching is too complicated / inertia 3.2

Decision of another person 2.8

More / better network coverage 0.8

Keep the same number 0.4

Other 4.0

Do not know / Did not respond 2.3

Source: ANACOM (2006).


consumers seemed focused on price as a driver of switching and even less

(about 4 percent) on quality of service.

Australia

The results of a questionnaire survey in Australia for ACMA (Australian

Communications and Media Authority – the telecommunications regula-

tor), summarized in Table 4.6, provide some additional information on

consumer satisfaction.

During the July 2005 to June 2006 period, satisfaction levels (defi ned

Table 4.6 Consumer satisfaction with telecommunications services in

Australia, 2004 to 2006

Respondents were asked: ‘In the last 12 months, how well have each of the

following service providers met your overall expectations?’ (Excludes ‘no answer’

and ‘doesn’t apply’)

Exceeded my

expectations

Mostly met my

expectations

Sometimes met

my expectations

Rarely met my

expectations

July

2004

– June

2005

(%)

July

2005

– June

2006

(%)

July

2004

– June

2005

(%)

July

2005

– June

2006

(%)

July

2004

– June

2005

(%)

July

2005

– June

2006

(%)

July

2004

– June

2005

(%)

July

2005

– June

2006

(%)

Local

telephone

company

7.8 8.2 73.9 73.3 13.3 13.5 5.0 4.9

STD

telephone

company

7.0 7.5 75.4 74.7 12.9 13.2 4.7 4.6

International

telephone

company

7.4 7.1 73.5 74.3 13.7 13.4 5.5 5.3

Mobile

phone

service

provider

9.4 9.9 72.5 72.1 13.1 13.4 5.0 4.6

Internet

service

provider

10.2 11.2 68.6 68.9 15.6 15.3 5.5 4.6

Source: Roy Morgan Research, questionnaire survey sample of approximately 23 000 people aged 14+ years, cited in ACMA (2006).


as ‘exceeded’ or ‘mostly met’ expectations) were more than 80 percent for

mobile telecommunications providers as well as fi xed line and Internet

service providers. The highest satisfaction levels were for long- distance

fi xed line telephone companies (82.2 percent). Internet service providers

were the providers who most exceeded expectations (11.2 percent).

Consumer expressions of the likelihood of them switching service pro-

vider off er another measure of consumer dissatisfaction with their current

provider. The Australian survey also provides some information in this

regard, summarized in Table 4.7. The likelihood of consumers switching

service provider ranged from: about 13.5 percent for fi xed line provid-

ers, and 14 percent for mobile telecommunications providers, to 16.5

percent for Internet service providers during July 2005 to June 2006. The

remaining consumers said they were either ‘unlikely’ or ‘neither likely nor

unlikely’ to switch providers.

Table 4.7 Consumer likelihood to switch providers in Australia, 2004–06

Respondents were asked: ‘How likely would you be to switch companies, if you

were able to buy that service from another company?’ (Excludes ‘no answer’ and

‘doesn’t apply’)

Very or fairly likely

to switch

Neither likely nor

unlikely to switch

Very or fairly

unlikely to switch

July

2004 –

June

2005

(%)

July

2005 –

June

2006

(%)

July

2004 –

June

2005

(%)

July

2005 –

June

2006

(%)

July

2004 –

June

2005

(%)

July

2005 –

June

2006

(%)

Local telephone

company

13.5 13.2 29.2 27.9 57.3 58.8

STD telephone

company

13.0 12.8 29.8 28.3 57.2 58.9

International

telephone

company

11.8 11.5 30.7 29.1 57.5 59.5

Mobile phone

service

provider

14.5 13.9 29.2 28.2 56.4 57.9

Internet service

provider

16.8 16.5 30.6 28.3 52.5 55.2

Source: Roy Morgan Research, questionnaire survey sample of approximately 23 000 people aged 14+ years. Cited in ACMA (2006).


United States

A report based on responses to a questionnaire survey of 1000 households

conducted in the United States in 2005 found that 36 percent of respond-

ents replied that early termination fees ranging from US$150 to US$240

(as shown in Table 4.8) had prevented them from switching (CALPIRG

Education Fund, 2005).

In fact 89 percent of mobile telecommunications consumers consid-

ered that early termination fees are designed to prevent consumers from

switching. They disagreed with the mobile telecommunications operators’

claim that the termination fees are a necessary part of the rate structure

(to enable them to recover the costs of subsidizing handsets, ‘buckets’ of

free calls, and so on).

4.3.3 Consumer Behavior in Internet Markets

United Kingdom

According to Ofcom’s research (2006b), awareness of alternative suppli-

ers is lowest in the Internet market with around 20 percent of Internet

consumers in the UK unable to name spontaneously any narrowband or

broadband Internet service providers (ISPs) in their area. However, half

of Internet subscribers were spontaneously aware of two or more nar-

rowband suppliers, and this rises to 74 percent when prompted (Ofcom,

2006b).

During 2003–06, just over one quarter (28 percent) of UK consumers

switched their ISP supplier. However, about three- quarters (72 percent)

had not switched during this time period and the majority (46 percent) had

not even considered doing so. Even among the 13 percent who had consid-

ered switching, not all had actively started looking for an alternative.

Table 4.8 Early termination fees of major US mobile telecommunications

service providers, July 2005

Company Early termination fee

Cingular US$150

Nextel US$200

Sprint US$150

T- Mobile US$200

Verizon US$175

Note: These operators control about 80% of the US mobile telecommunications market.

Source: CALPIRG Education Fund (2005).


The fi gures in Table 4.9 indicate that only about 10 percent of broad-

band consumers in the UK switched suppliers during 2007 and 2008,

although about another 20 percent had considered doing so. Another 8

percent had considered switching, but had not actively started looking for

an alternative. However, more than 60 percent had not even considered

doing so.

Ofcom found that regardless of whether or not they had switched ISP

in the last four years, almost three- quarters (72 percent) had made some

change to their existing service with their current ISP in this time period.

This was mainly the result of consumers switching connection type (62

percent had done this) and/or switching tariff , including upgrades to

connection speed (53 percent had done this). A few (10 percent) had

renegotiated their deal (that is, asked their current ISP to match a better

deal they had seen elsewhere). Thus, even though the research shows that

many consumers in the Internet market have not considered changing

their ISP, competition had allowed them to ‘participate’ in other ways.

Compared with broadband users, narrowband users were signifi cantly

less likely to have made any changes to their existing package: 24 percent

had changed connection type and 25 percent had changed their tariff or

package.

According to Ofcom’s research (2006b), the greatest deterrents to active

participation in the Internet market include the possibility of getting

locked into a contract with a new supplier (68 percent of consumers);

reluctance to leave a known and trusted ISP for one that was unfamiliar

(63 percent of consumers); and the perceived eff ort of shopping around,

including diffi culty in making comparisons between ISPs (44 percent)

and not enough time to research all the options (40 percent). The main

drivers of participation include interest in technology, and desire for low

Table 4.9 Switching in the UK broadband communications market during

2007 and 2008

2007 (%) 2008 (%)





None 65 62

100 100



cost and/or willingness to consider unfamiliar brands if off ered a good

deal. There were few demographic diff erences, suggesting participation is

infl uenced by attitudes and behavioral factors rather than by diff erences

in demographics.

This information suggests that in the face of complexity (perceived or

actual), inactive consumers are adopting fallback, risk- averse strategies

and will stay with what they know, even though it might not be the best

option. The information is consistent with the argument of behavioral

economics that an endowment factor serves to infl uence decisions in favor

of the present provider.

4.4 IMPLICATIONS FOR POLICY AND REGULATION

4.4.1 Regulation of Information Disclosure

More information is usually seen to be desirable. A demand- side behav-

ioral perspective warns that if consumers have limited cognitive abilities,

either generally or in a particular situation, then adding more information

may result in information overload and hence in worse decision- making

(Camerer, 2003). Excessive disclosure can confuse consumers (as evi-

denced in the case of mobile phone and Internet tariff options) and can

also discourage fi rms from providing useful information through their

advertising. In these cases, the need is not for more but for better (perhaps

less) information in a structured, easily comprehensible format (Gans,

2005a).

Typically, a mandated disclosure is intended to improve the informa-

tion received by the less- informed buyers to a greater extent than by those

more sophisticated consumers who may already possess and be able to

process the requisite information. But the reverse can also be the case

where it is the more sophisticated (and more cognitively capable) consum-

ers who can make more use of the information contained in a mandated

disclosure. For example, mandating information on broadband download

speeds and caps can be of far more use to those consumers who have at

least a basic knowledge of technical features of broadband capacity (for

example, bits per second).

Diff erences in the incidence of benefi ts from mandated disclosure means

that any costs could tend to be diff erent as well. Some consumers may be

provided with information they cannot use, some with information they

do not need, and some others will be overloaded with information, but all

will have to pay for the compliance and related costs (OECD, 2006).


This cautionary note aside, there are many situations where an informa-

tion disclosure remedy is necessary and appropriate. Even in competitive

markets, suppliers may possess market power if consumers are not well

informed about products, supply conditions and/or alternatives and feel

unable or unwilling to switch between these alternatives (ERG, 2009).

Providing consumers with more and better information, for example

through mandatory disclosure or through third party certifi cation, may

facilitate more active participation in the market and wiser decision-

making (Better Regulation Executive, 2007).

Just as informational problems may be multifaceted, so too may

problems that result from behavioral biases. It is important for policy

and regulation to recognize these biases and develop a fuller under-

standing of the needs and motivations underlying consumer behavior

in telecommunications markets. Not all behavioral biases lead to

consumer detriment. For the most part they do not lead consumers to

depart signifi cantly from optimal decisions. Public policy should be con-

cerned only with those biases that lead to signifi cant detriment. When

markets fail because of such costly biases, remedies should be shaped

accordingly. For example, a situation of choice or information overload

could be aggravated by a requirement for more information disclosure.

Rather, the appropriate intervention may involve reframing the infor-

mation that is available to consumers in a way that makes choice easier

(OECD, 2007).

The policy implications of demand- side analysis in some respects are

more challenging than those derived from the more traditional approach

to consumer protection and empowerment. For instance, behavioral eco-

nomics predicts that for various reasons some consumers (or consumers

in some circumstances) may act in ways that are inconsistent with their ex

ante preferences. Consumers may use information in ways not predicted

by neoclassical theory or they may, for various reasons, not use available

information. Thus, while in some cases providing more information or

providing information in a diff erent form may remove or reduce the risk

to consumers, this will not always be the case. If it is to be eff ective, an

eff ort to inform consumers must appreciate how people actually think.

The behavioral response to identical pieces of information will depend on

how they are presented and framed.

Thus, in circumstances where the conduct of suppliers alters the pref-

erence set of consumers and hence their choices, resulting in an inferior

outcome for those consumers, the solution may lie in regulatory interven-

tion that aims ‘to steer people’s choices in welfare- promoting directions

without eliminating freedom of choice’ (OECD, 2006).


Resetting defaults

Options can be presented in ways that lead the consumer to gravitate

towards certain choices that are in their interests. For example, to over-

come the biases of hyperbolic discounting and of default inertia, renewal

of a telecommunications contract can be presented with ‘opting out’ as

a default. Policy- makers and regulators could consider the use of ‘light’

interventions such as a mandatory ‘opt- in’ default provision (rather than

an opt- out default) as part of arrangements for extending a telecommuni-

cations contract.

Reframing

Suppliers can be required to present information in a variety of frames, or

in specifi c frames that may guide sound consumer choice. The policy solu-

tion under a demand- side behavioral approach is that rather than requir-

ing that the consumer read ever more complex contracts, the government

can mandate standard form clauses or even standard form contracts. This

recognizes that consumers probably will not read the contract in detail. But

care should be exercised so that in removing ambiguity, the behavioral solu-

tion also tries not to limit consumer choice, since this can make consumers

worse off if parties are now constrained to government- devised contracts

that are infl exible and cannot be altered to fi t their personal circumstances.

General de- biasing

Although biases are generally deeply ingrained, there is a role for con-

sumer education, making consumers aware of their biases, and helping

them to develop mechanisms to overcome those biases that go against

their self- interest. For instance, to counter ‘hyperbolic discounting’,

consumers might be counseled to consider carefully whether a fl at rate

contract or a ‘free’ mobile phone handset with a two- year contract really

does suit their usage level and pattern. Cooling- off periods can be used to

allow consumers to reframe their choices and to give them an opportunity

for rational reconsideration to overcome the infl uence of impulsive choice,

such as those resulting from hyperbolic discounting.

Recognizing risk- averseness in the switching process

In addition to making more information available, raising awareness, and

addressing some of the behavioral biases that may prevent consumers

from actively participating in the market and making decisions to switch

where not satisfi ed, there are some specifi c measures that can be taken

by regulators to reduce practical impediments to consumers switching

from one supplier to another. For example, regulators could ensure that

the shortest possible time is taken to complete number portability for


consumers switching fi xed line and mobile telecommunications provid-

ers. Regulators could require that all Internet service providers ensure a

simple, costless (or at least cheap) and quick transfer of consumers who

choose to switch provider. Regulators could examine the need to limit the

‘lock- in’ period for mobile phone handsets in order to facilitate switching.

Where applicable, the fee for unlocking the handset should be related to

the cost involved.

Vulnerable consumers

The circumstances of vulnerable consumers warrant special attention. For

instance, Ofcom’s research in the UK indicates that while those without

Internet access – frequently older and low- income consumers – are less

likely to look for information at all, providing price information solely

via websites could risk excluding relatively large and vulnerable groups of

consumers. There is need for further research to assess:

Consumers’ use of information sources and how consumers use ●

these when making choices.

Consumer opinion on current information obtained through, for ●

example, supplier websites, the Internet generally and specialist

publications.

The importance of savings in infl uencing decisions to switch sup- ●

plier, what trade- off s are made when considering whether to switch,

and the key drivers in consumer decisions to search or switch.

The extent to which misperception, low awareness of achiev- ●

able savings or ease of the switching process impacts on consumer

switching decisions.

There is particular need to assess the needs and motivations of consumers

– especially uninvolved and vulnerable consumers – in more detail to

ascertain what, if anything, would encourage more participation in tel-

ecommunications markets. This would assist consideration of whether,

and if so what, regulatory intervention is warranted. All consumers want

simply understood information. If businesses and policy- makers design

the regulated information with vulnerable consumers in mind, all consum-

ers are likely to benefi t.

4.5 CONCLUSION

This chapter distils some basic elements of behavioral economics and

demand analysis, examines available evidence of actual consumer behavior


and considers implications for telecommunications policy and regulation.

The chapter concludes that the insights of a demand- side perspective,

including behavioral economics, can serve to alert policy- makers and

regulators in the telecommunications sector of the need for a number of

measures:

Policy- makers and regulators should develop a better and fuller ●

understanding of the needs and motivations underlying consumer

behavior in telecommunications markets, especially those of vulner-

able consumers (such as those in rural areas, the elderly, minors,

disabled, those on low incomes, the unemployed).

Policy- makers and regulators could assist consumer participation ●

in telecommunications markets by educating consumers about

their rights, by raising awareness about new services and options

off ered by the market, and by making the process of switching in

the fi xed line, mobile and Internet markets easier, cheaper and

faster.

Regulators should consider requiring that all major operators ●

provide complete, comparable, appropriate and accurate informa-

tion to consumers through diff erent channels (for example, through

leafl ets, radio, consumer hotline and web- based programs) to enable

consumers, especially vulnerable consumers, to identify quickly

the most suitable and best- value telecommunications plan (ERG,

2009).

Regulators could use more eff ective means of targeting information ●

to vulnerable groups to provide them with practical guidance about

how they can get the best deal.

Regulators could encourage third parties, including consumer ●

organizations, to provide price and service comparison facilities

through consumer hotlines, websites, and so on.

Regulators could work with fi xed line (including Internet service ●

providers) and mobile network operators to develop and publicize a

set of comparable indicators relating to quality of service.

Regulators should ensure that the shortest possible time is taken to ●

complete number portability for consumers switching between fi xed

line and between mobile service providers.

Regulators should require that all Internet service providers ensure ●

a simple, free (or at least low- cost) and quick transfer of consumers

who choose to switch provider.

Where not already in place, regulators could require truth- in- billing, ●

and restrict harmful business conduct and practices (for example, by

prohibiting mis- selling and misleading advertising).


REFERENCES

ANACOM (2006). Electronic communications consumer survey, February 2006 – key fi ndings. May.

Australian Communications and Media Authority (2006). Telecommunications performance report 2005–2006. October.

Barrow, P. (2007). Just enough: empowering fi xed- line telecommunications con-sumers through a quality of service information system. CCP Working Paper 07- 2. Retrieved from http://www.ccp.uea.ac.uk/publicfi les/workingpapers/CCP07- 2.pdf.

Better Regulation Executive (2007). Government response to the fi nal better regu-lation executive/National Consumer Council report on consumer information. UK Department for Business Enterprise and Regulatory Reform, London.

CALPIRG (California Public Interest Research Group) Education Fund (2005). Locked in a cell: how cell phone early termination fees hurt consumers.

Camerer, C. (2003). Regulation for conservatives: behavioural economics and the case for asymmetric paternalism. University of Pennsylvania Law Review, 151, 1211–54.

ERG (2009). Report on transparency of tariff information.Gans, J.S. (2005a). Real consumers and telco choice: the road to confusopoly.

Paper presented to the Australian Telecommunications Summit, Sydney. 21 November.

Gans, J.S. (2005b). Protecting consumers by protecting competition: does behav-ioural economics support this contention? Competition and Consumer Law Journal, July, 1–11.

Miravete, E. (2007). The doubtful profi tability of foggy pricing. Retrieved from http://www.eco.utexas.edu/facstaff /Miravete/papers/EJM- Foggy.pdf.

NERA (National Economic Research Associates) (2003). Switching costs. A report prepared for the Offi ce of Fair Trading and the Department of Trade and Industry. Economic Discussion Paper 5. April.

OECD (2006). Roundtable on demand- side economics for consumer policy: sum-mary report. Paris. Available at http://www.oecd.org/dataoecd/31/46/36581073.pdf.

OECD (2007). Summary report on the second OECD roundtable on consumer policy. Paris. DSTI/CP (2007)1.

OECD (2008). Enhancing competition in telecommunications: protect-ing and empowering consumers. DSTI- ICCP- CISP (2007)1/FINAL, May, 2008, Retrieved from http://www.olis.oecd.org/olis/2007doc.nsf/LinkTo/NT00005FB2/$FILE/JT03246386.PDF.

Ofcom (2006a). Ofcom’s consumer policy – a consultation. Retrieved 8 February 2006 from http://www.ofcom.org/consult/condocs/ocp/ocp_web.pdf.

Ofcom (2006b). The communications market 2006.Ofcom (2008). The consumer experience 2008. Telecoms, internet and digital

broadcasting. 24 November.Pesendorfer, W. (2006). Behavioural economics comes of age: a review essay

on advances in behavioural economics. Journal of Economics Literature, 44 (September), 712–21.

Sylvan, L. (2006a). The interface between consumer policy and competition policy. Department of Consumer Aff airs, Victoria. Lecture in honour of Professor Maureen Brunt.


Sylvan, L. (2006b). Capturing the consumer interest. Consumer Panel Toolkit launch. Australian Competition & Consumer Commission. 2 February.

UK National Consumers Council (2006). Switched on to switching? A survey of consumer behaviour and attitudes, 2000–2005. April.

Waddams- Price, C. and C. Wilson (2005). Irrationality in consumer switch-ing decisions: when more fi rms may mean less consumer benefi t. Industrial Organization working paper from EconWPA.

PART II

Technology convergence and the future role of competition and regulation

109

5. The measure and regulation of competition in telecommunications markets1

Marcel Boyer

5.1 INTRODUCTION

Telecommunications and more generally information and communication

technologies (ICTs) constitute the backbone of our societies. Nothing is

more important for the well- being, present and future, of modern societies

than getting an effi cient web of telecommunications chains and networks.

Such a web is essential for social cohesion, productivity gains, innovation

and commercialization, and for reaching the highest level of humanist

economic growth.

The development of a telecommunications web is signifi cantly infl u-

enced by the regulatory framework put in place to oversee the evolution of

this web towards a competitive system through policies aimed to protect

effi cient newcomers from predatory incumbents and to protect the public

against the capacity of large fi rms to exercise market power. If and when

the level of competition is deemed suffi cient, specifi c economic regulation

is expected to disappear in favour of general competition or antitrust

policy while technical, architectural or design regulation and coordination

will subsist.

When is the level of competition in a national telecommunications

industry suffi cient? When it is not, then when will it be: that is, what value

of which indicators will tell us that its time has come? The answers to those

questions are vital for any society since too much economic regulation for

too long as well as too little economic regulation for too short a time gen-

erate important social costs in terms of consumer welfare, productivity,

innovation and growth.

The specifi c objectives of this chapter are, fi rst, to develop a methodo-

logical framework, which will allow a proper characterization of the level

of competition in the residential local access market; and second, to rec-

ommend a new approach to regulation. It addresses major concerns of the


regulators regarding two ‘factual measurements’: fi rst, the small number

of competitors in the residential local access; and second, the high market

shares captured by the incumbent fi rms, even after a decade of favourable

policies aimed at ‘more competition’ in that market. The situation is basi-

cally the same in all regional residential local access markets.

The above ‘factual measurements’ are poor indicators of the level of

competitive pressures in those markets. Indeed, those facts may be cred-

ible indicators that competitive pressures are relatively intense, so much

so that entry strategies by newcomers are not profi table. The traditional

measures of competition based on market shares are inadequate because

this industry has more characteristics of an emerging industry than of a

mature industry.

The current approach to economic regulation, mainly based on the tight

control of incumbent local exchange carriers (ILECs) and light- handed

surveillance of competitive local exchange carriers (CLECs), has run its

course as market delineations become rapidly blurred.

The new regulatory framework should rest on three specifi c principles

of economic effi ciency:

The pursuit of a dynamic regulatory approach based on implement- ●

ing proper competition processes and information systems.

The promotion of competition through proper incentives ensuring: ●

– dynamically effi cient inter- access prices and conditions;

– effi cient investment programs in network maintenance and

development.

The design of non- predatory pricing rules through full cost sharing, ●

to promote the emergence of a more competitive industry, even if

such rules reduce static effi ciency.

If the government or regulator wishes to adopt a proactive strategy

to favour the emergence of competition, based on the fact or belief that

incumbents have an ‘unfair’ advantage, then the preferred policy is a

direct incentive subsidy, such as a generous but deferred investment grant

or tax credit to be paid at some time in the future conditional on the

entrant’s capture of a predetermined market share. Only those potential

entrants with superior products or services, and/or superior technology,

and/or better- quality consumer service, who believe they can compete

with and displace the incumbent one way or another, will enter or con-

sider entering the industry. Ineffi cient competitors, fl y- by- night operators

and fast- buck seekers will stay out. Consumers and customers will reap

the benefi ts of effi ciency- enhancing competition through proper creative

destruction.

The measure and regulation of competition 111

5.2 THE RELEVANT MARKET

The telecommunications industry looked very much like a mature indus-

try in the 1970s and even in the 1980s. But over the last 20 years the devel-

opment of new information and communication technologies made the

telecommunications industry appear much more like an industry in con-

stant instability, with new technologies, new products, new competitors,

signifi cant level of mergers and acquisitions as well as divestiture activity.

These movements were accompanied by signifi cant changes in regulation,

from earnings- based or rate- of- return- based cost plus regulation towards

a whole array of incentive, price cap, light- handed, competition- based and

competition- enhancing regulation.

Regulators in most if not all jurisdictions have forced incumbent fi rms

who own the local loop (the last mile) of the wireline network to give access

to their network to CLECs at relatively favourable conditions. In spite of

such eff orts, the level of access eff ectively demanded and used remains quite

low, raising fears that ILECs may have been able, for some unknown reasons

and through some obscure means, to prevent and/or block access, thereby

maintaining their market power over the local wireline access services.

However, market shares (whatever way they are measured in practice)

may not be the best measure of competitive pressures in a fast- growing

and technology- driven industry. The actual and potential competitive

products and services in the telecommunications industry are in a constant

state of fl ux. Besides the strictly regulated products and services, we see

changes in products and services, appearing in a seemingly unpredictable

manner, as well as changes in the prices, price structures (fi xed or access

charge, variable charge, bundled services discounts, and so on), restricted

conditions imposed on diff erent groups of consumers of those products

and services, and even their technology delivery platforms, aff ecting their

accessibility and fl exibility of access.

The relevant actual market is therefore diffi cult to determine and the

empirical studies that can be conducted can only be incomplete as to

what is the real level of competition in the industry. Even the best state-

of- the- art econometric studies available are subject not only to explicit

caveats but also to a signifi cant level of cavilling given the simplifi cations

and aggregations in product characteristics and in pricing structures that

researchers and analysts must inevitably have recourse to in their empiri-

cal estimation of demand structures for telecommunications products and

services. One must treat these attempts very carefully and with skepticism.

Even if they may nevertheless contain some useful information, the use

one may make of the results in order to fi ne- tune the regulatory frame-

work remains very limited.


In most cases, consumers face a menu of pricing structures and it

becomes diffi cult, in empirical statistical or econometric studies, to sum-

marize such pricing structures in a meaningful notion of what is the ‘price’

of a product or service to be compared to the price of other (substitute)

products or services. There are many reasons why telecommunications

products and services are priced in such a way.

Menus of pricing structures or calling plans, according to which con-

sumers pay a fi xed charge per unit of time (typically a month) and then

decreasing or increasing price levels for increasing consumption levels, are

or can be implemented when the resale of the product can be controlled or

deterred in such a way that the end consumer of the product can be iden-

tifi ed by the producer or seller. Menus of contracts or calling plans have

three important impacts: fi rst, they promote effi ciency and favor a faster

development of the industry; second, they generate higher profi ts through

price discrimination; and third, they allow product diff erentiation thereby

softening competitive pressures.

Menus of pricing and contracts promote effi ciency insofar as the mar-

ginal (last) units of service are sold at or close to marginal cost. Hence at

the margin the most important allocative effi ciency condition (marginal

cost pricing, possibly equal to zero) is satisfi ed. The welfare value that

consumers attach to those marginal units is equal to the cost of produc-

ing or generating those marginal units. Hence the effi cient social- welfare-

maximizing level of production and consumption is achieved; at least once

the decision to subscribe is made. However, a surplus may be lost if some

of those potential consumers do not subscribe to the service because of the

fi xed subscription cost, whether a one- time cost or a monthly cost.

Carefully determining the calling plan limits and conditions allows tel-

ecommunications fi rms to diff erentiate their products and services from

what competitors are off ering, thereby softening the level of competitive

pressures across the industry. Off ering the same or very similar calling

plans as those off ered by another competitor (same limits, conditions and

characteristics) would force the two fi rms to compete on prices only. This

latter type of competition is likely to end up in a money- losing price war.

Many factors other than pricing drive the demand(s) for telecommu-

nications products and services. To avoid competing strictly on prices,

fi rms will compete along those other factors: coverage (and dead zones),

digital versus analog transmission, physical and design features of phones

(such as color, size and weight), battery life, digital subscriber line (DSL)

services and cable modem services, household composition, education,

employment and occupation, security, commuting or not, and so on.

Consumers are most likely indiff erent between the diff erent forms or

technologies by which they satisfy their needs for telecommunications


services, as long as their demand for connectivity and fl exibility, the

two most important characteristics that they value, are comparable in

terms of quality and aff ordability. In most cases, consumers are not

even aware of the technological characteristics of the platform on which

their calls and other services are transmitted. It is more the end product

that counts: connectivity, fl exibility, safety, dependability, accessibility,

capacity (high speed and broadband) and user- friendliness. In that sense,

the demand expressed by consumers for diff erent telecommunications

devices (wireline, wireless, cellular mobile, satellite- based mobile, Internet

Protocol – IP – telephony, and so on) is a derived demand rather than a

direct demand. Connectivity refers to the geographic area (or to the set of

potential called and calling parties) over which communications can be

established; connectivity is a multidimensional characteristic since it often

relies on a multilayer pricing plan: diff erent connectivity at diff erent prices.

Flexibility refers to the availability of the service under diff erent circum-

stances, the most important certainly being the availability throughout a

given geographic area, as one moves around in the area. Safety refers to

the health hazard one may be exposed to in using the service on a regular

basis, an example being the concerns for safety which have prompted dif-

ferent jurisdictions to forbid the use while driving of handheld cellphones,

and in some cases all uses of cellphones. Dependability refers to the assur-

ance of service of a high quality level, for example free of parasites and

free of breach of confi dentiality, when one needs it, especially but not

only in an emergency situation. Accessibility refers to the availability of

the service when one wants to communicate with another party. Capacity

refers to the possibility to transmit vast amounts of information, such as

large fi les, and high- defi nition pictures as well as videos, at a rate high

enough to quasi- replicate in situ communications. User- friendliness refers

to the ease of use of the technology and service.

The empirical estimates of the level of substitutability (and therefore

competition) between wireline and wireless telephony is in terms of the

own- and cross- price elasticities of demands for those services. To run

such empirical studies, one needs suffi cient variability throughout the

sample observations. If every observation unit in the empirical sample

subscribes to a wireline service, then it is impossible to identify the decision

variables behind or justifying the subscription itself. To do so, one needs

within the sample a reasonable number of non- subscribers to the diff er-

ent wireline and wireless services, a condition which may not always be

satisfi ed for the basic wireline services. However, it may still be possible to

estimate empirically the relative importance of the explanatory variables,

drivers or shifters, of the usage level (like minutes) of the services.

In terms of data requirements, it is clearly preferable to have household-


level data rather than more aggregated regional data. A study2 by Rodini

et al. (2002) of wireline–wireless communications substitution behavior

using household- level data reports own- price elasticities of mobile access

demand and cross- price elasticities with relatively high degrees of preci-

sion. The results suggest that wireline and wireless telephony may be, for

a signifi cant number of consumers, complements rather than substitutes,

although the negative (net) cross- price elasticities indicate that on average

the substitution eff ect dominates the complementary eff ect. The devel-

opment of a larger number of diff erent telecommunications technolo-

gies, each with important specifi c characteristics (connectivity, fl exibility,

safety, dependability, accessibility, capacity, friendliness), tends to increase

the global usage level of all telecommunications technologies. Computing

market shares in terms of telephone lines is not the only way to measure

concentration. In fact, it now appears to be an outdated way to do it.

5.3 THE LEVEL OF COMPETITION

Let us fi rst consider the level of substitutability in the local telephone

market between wireline and wireless communications. The standard

approach of antitrust economics is to consider the level of competition

as suffi cient to prevent an undue exercise of market power by incumbents

if those incumbents would fi nd it unprofi table to implement a small but

signifi cant (typically 5 percent) and non- transitory (typically for one year)

increase in price (SSNIP) for their services, given the pricing structure of

their competitors’ products and services. Such a move would be profi table

if the reaction of consumers to such a price increase for one year were rela-

tively small and limited, that is, if there were no close substitutes to which

they can turn to off set the impact of that price increase.

To do so, one must defi ne the relevant market, the relevant choice set

that consumers face regarding the satisfaction of their telecommunica-

tions needs, and the relevant set of competitors to determine the level of

competition in the telecommunications industry. These questions can be

raised in the study of any industry or industrial sector but are particularly

important in the context of the telecommunications industry, which has

many characteristics of an industry in its early phases of development

rather than a mature industry, a rather striking example of the ‘back to

the future’ storyline.

The ever- increasing number of consumers having access to both tech-

nologies and having access soon to a third one, the Voice over Internet

Protocol (VoIP) technology, continuously reduces the need for protec-

tion. The level of market power of the ILEC is seriously diminished up to


a point where more regulation may cause more harm than good. Insofar

as those consumers who already have the possibility to react to any price

increase by switching to alternative modes of telephone communications

cannot be isolated from other consumers – that is, cannot be off ered a

specifi c pricing plan – the wireline incumbent will have to balance the

benefi ts from an increase in the price with the loss of revenues due to such

switching. Hence, on that basis alone, one would be inclined to recom-

mend immediate deregulation even if some consumers have, by choice or

not, access to wireline local telephony only.

The potential future competition in products and technologies also plays

an important role and defi ning the set of competitors, actual and potential,

is a critical step in determining the level of competition. The identity and

role of potential competitors is diffi cult to determine since many products

and services as well as some future providers may simply not be yet identi-

fi ed even though they may be dans les coulisses and certainly present in the

minds of incumbents, whether they are wireline, wireless or VoIP provid-

ers. Any empirical study of the type econometricians usually do cannot

avoid neglecting such crucial aspects.

Moreover, telecommunications products and services must be seen as a

‘technology’ to generate a vector of communications (and entertainment)

characteristics that customers value. A consumer is likely to express his

or her demand for a given product or service in terms of its contribution

to the overall connectivity, fl exibility, safety, dependability, accessibility,

capacity and friendliness of his or her portfolio of products and services,

which are off ered or distributed through diff erent technologies (wireline,

Internet access, many mobile lines, and so on). It is not the specifi c stand-

alone characteristics of any given service that are likely to count, but their

contribution to the individual communications portfolio. This has impor-

tant implications for our understanding of the proper regulatory approach

to competition in telecommunications, in particular in the local wireline

access. A product- by- product or market- by- market approach may turn out

to be very detrimental to the welfare of consumers because it is likely to miss

the interdependencies between products and services and the specifi c con-

tribution of a product or service to the group or portfolio of telecommuni-

cations means, products and services that a customer may have access to.

5.4 THE COMPETITIVE ENVIRONMENT

One must look at local wireline access competition from a ‘process’ point

of view rather than from a ‘market share’ point of view. This makes the

analysis slightly more diffi cult and complicated but at the same time more


transparent, more adequate, and better fi ne- tuned to the characteristics of

the telecommunications industry.

The process approach concentrates on ensuring open access to the exist-

ing network facility at properly defi ned competitive access pricing and

conditions, rather than on the number of fi rms demanding access or the

market shares of those fi rms as compared with the incumbents’ market

share. In so doing, it avoids opening the doors to ineffi cient competitors

who may benefi t from the regulators’ overemphasis on market shares at

the expense of consumer welfare.

If not properly addressed, these developments may turn out to be value-

destroying by reducing the incentives to invest in developing and maintain-

ing infrastructure capacities that are prone to social inclusion, in particular

in terms of the obligation to serve. The situation is even more dangerous

if potential competitors are in fact unwilling to enter the (residential) local

wireline access market at any cost. We will see next that there are good

reasons to suspect that this may indeed be the case, and that the regulators

are ill advised to concentrate their attention on market shares rather than

on the competitive process. The process approach to fostering competition

is also likely to ensure a level playing fi eld for all competing players while

at the same time recognizing the diff erent responsibilities of those players,

especially in developing and maintaining infrastructure capacities.

It is important to understand the competition model relevant for the tel-

ecommunications sector, including the pro- competitive or anticompetitive

role of the following factors: the number of competitors, the substitutabil-

ity between the diff erent products and technologies, and the existence of

excess capacity. Equally important is the understanding of the entry and

exit strategies of competitors (based on ‘ex post entry’ market conditions

rather than on the ‘actual ex ante entry’ conditions) as well as the dynamic

incentives for research and development (R&D) and innovation. Let us

consider these in sequence.

Consider an industry with the following three characteristics: fi rms are

engaged predominantly in short- run competition, the products of the dif-

ferent fi rms are close quasi- homogenous substitutes, the fi rms have no

binding capacity constraint and can produce up to the market demand at

a common quasi- constant marginal cost (with some fi xed costs). In such

an industry, a Bertrand industrial equilibrium will most likely emerge

(Boyer et al., 2004b; Boyer et al., 2007). In such an equilibrium, the com-

petition between the fi rms, which are eager to increase their market share

by pricing their product slightly below their competitors’ prices, will make

prices fall towards the marginal cost level and therefore generate negative

profi t levels given by the negative of the fi xed costs. Such a situation is

of course extreme but it is nevertheless quite instructive. It shows that a


strategy to enter into an industry characterized by product homogeneity,

constant marginal costs (excess capacity) and short- run competition is

likely to be a money- losing strategy. Hence, entry in such an industry is

very unlikely. Entry by competitors in such an industry cannot be expected

unless those Bertrand competitive conditions can be changed or bypassed

and controlled in some ways. Hence the importance for the competitors

to plan their entry in the industry with a strategy aimed at avoiding the

Bertrand competition conditions.

A fi rm could profi tably enter with a diff erentiated product (possibly and

preferably off ered through a diff erent technological platform) in terms of

those characteristics valued by the consumers, namely connectivity, fl ex-

ibility, safety, dependability, accessibility, capacity and user- friendliness.

Given the diversity of consumers’ tastes and needs, there may be a niche

in which the fi rm can enter and make profi ts. A fi rm could also profi tably

enter into the industry with a limited capacity (using a ‘judo’ economics

strategy) such that the incumbent will fi nd it unprofi table to engage in a

price war in order to prevent the small and limited loss of customers that

the capacity- constrained entrant can attract. The telecommunications

frenzy of the 1990s, in terms of both capacity investments and technologi-

cal developments, led to a 500- fold transmission capacity increase while

demand itself increased by a phenomenal fourfold. It is estimated that $200

billion of telecommunications network capacity was built unnecessarily (ex

post). This building of overcapacity in a ‘commodity’ market brought the

telecommunications industry into the Bertrand competition trap.3

A fi rm could also profi tably enter with a set of products and services

relatively similar and homogenous to that of the incumbent and with a

relatively large capacity, thereby credibly signalling to the incumbent that

it is entering for the long term. If the entrant’s strategy is suffi ciently con-

vincing, the incumbent may fi nd more reasonable and profi table to respond

to entry by adopting an accommodation strategy in order to avoid a likely

unsuccessful but potentially very costly price war. In the repeated competi-

tion game, the fi rms may be able to fi nd mutually benefi cial accommoda-

tion strategies avoiding the Bertrand competition outcome.4

But entering with such a strategy is quite risky given that the incumbent

may not be convinced that it is in its best interest to accommodate entry.

Inducing an entry- accommodating reaction by the incumbent will in

general require that the entrant either incurs signifi cant investment outlays

as sunk irreversible investment costs, or signs contracts with important

customers with conditions that are very costly for the provider for rene-

gotiating or breaking the contract, or adopts a bridge- burning (no- escape)

strategy to signal clearly and convincingly to the incumbent its forceful

intention to remain in the industry (by raising its cost of exiting).


That is how we must understand the competition for the residential

local wireline access and why entry in this market has been so weak even

after multiple pro- entry decisions and actions by telecommunications reg-

ulators since 1997. The residential local wireline market is most likely to be

characterized by Bertrand competition conditions: short- run competition

(favoured by the CLEC immunity regarding the development and main-

tenance of the network capability, which makes entry into the industry

almost as easy as exit from the industry), homogenous product or service

(there is arguably nothing more homogenous in the telecommunications

industry than the local wireline services off ered by two diff erent provid-

ers), constant marginal cost (with this marginal cost being very low), and

potentially no binding capacity constraint (by getting access to the whole

local wireline network at low cost).

Entry into the local wireline market is clearly not an easy policy to

pursue and could potentially be a very risky strategy. One may expect that

entrants will prefer to invest in other segments of the telecommunications

industry. However, this is not peculiar to the telecommunication industry

and markets. Similar diffi culties and risks are present in most industries.

Indeed, incurring losses during the early phase of entry is quite common

for most if not all entrants in any industry. Such a situation would not

justify a regulatory intervention to protect entrants from the competitive

pressures they will be confronted with. When Toyota decided in the 1950s

to challenge the major incumbents of the time in the automobile industry,

it came in with a superior technology and eventually a better product.

There were no calls for protection from the incumbents through price and

marketing controls (except for the legal provisions forbidding predatory

pricing). The company made the risky gamble that its value chain model

was better, that is, would lead to lower costs and higher quality, than the

value chain model of the big three of the day. And that is the way competi-

tion serves customers through creative destruction. In the telecommuni-

cation industry, such creative destruction could come from incumbents

entering into each other’s markets or from new entrants capitalizing on

their expertise in a related industry (cable or satellite communication serv-

ices for instance). Many other examples could easily be given.

5.5 A RENEWED PRO- COMPETITION REGULATORY FRAMEWORK

The real conundrum of the regulator is to create and enforce static com-

petitive conditions between the current bundles of products and services

or technologies as well as dynamic competitive conditions yielding strong


intensity of incentives for developing new technological platforms, intro-

ducing new products and services that could be both cheaper and more

effi cient, connectivity- wise and fl exibility- wise, and in so doing replace

the current goods and services and possibly their producers and distribu-

tors. To achieve a proper balance between short- run (static) and long- run

(dynamic) goals, regulators must rely on competitive processes, that is,

they must make sure that their interventions are not aimed at micro-

managing prices and quantities but rather at making sure that those prices

and quantities emerge from a competitive environment. Micro- managing

prices and quantities is very much reminiscent of the old regulatory frame-

work (rate- of- return regulation for instance) that led the regulator to

discover through adversarial proceedings the ‘true’ cost functions of the

fi rms. We know now that this objective was futile: economic costs are not

accounting costs. Exerting a proper level of competitive pressures is the

only way to induce the fi rms really to minimize the costs of their portfolio

of goods and services, that is, to make the accounting costs close or equal

to the economic costs.

The competition processes view of regulation developed below stresses

the importance of competition as the most powerful and effi cient generator

of social effi ciency in the telecommunications industry. Benevolent regula-

tion can, under complete and perfect information, favor the emergence of

an effi cient allocation of resources in a partially natural monopoly indus-

try such as telecommunications (proper goods and services, proper prices,

proper investments in R&D, capacity building and network maintenance,

and so on) and in so doing can dominate competition. But in an incom-

plete information (on technologies, consumers’ preferences and needs,

and R&D and investment opportunities) situation such as the one prevail-

ing in the telecommunications industry, it is these competitive processes

and pressures that can generate the best feasible allocation. The proper

reference point here is not the full- information fi rst- best allocation, but

rather the incomplete- information second- best allocation. Many confl icts

regarding the regulators’ role, ways and means stem from the ill- conceived

comparison between an infeasible fi rst- best allocation and a feasible

second- best allocation that takes into account explicitly the informational

constraints that the benevolent regulator is facing: the (fi rst- )best is here

clearly the enemy of the good.

The possibility and desirability to inform customers better about the

characteristics, including pricing, of diff erent (packages of) telecommuni-

cations goods and services should, in the context of an emerging industry

such as telecommunications, be a well- understood role of the regulator.

This regulation- by- information approach is not new but it is too often

neglected as if customers of complex goods and services were individually


able to make and understand comparisons of pricing and other charac-

teristics such as connectivity, fl exibility, safety, dependability, accessibil-

ity, capacity and user- friendliness of telecommunications services. In an

industry such as telecommunications, the generation and presentation of

such information in a user- friendly way should be a role and responsibil-

ity of the regulator, at least till the customers develop such a capability

of their own. Asking the regulator to assume such a responsibility could

prove very important for effi ciency and political acceptability during the

transitory period towards a complete deregulation of the telecommunica-

tions industry.

Regulators generally recognize that substituting competition mecha-

nisms for costly imperfect regulation could improve cost- effi ciency as well

as allocative effi ciency. But for network industries, it is not necessarily

desirable to introduce competition in all segments. Some segments of a

network, which are essential inputs for the potentially competitive seg-

ments, may be subject to signifi cant economies of scale. These so- called

essential facilities, specifi c to a given technology, could remain price

regulated as long as alternative technologies allowing those segments to

be bypassed are not available. For instance, the local wireline may still

represent a non- standard natural monopoly. Even though it is challenged

by competing technologies, it would not make sense to duplicate the last

mile wireline connection to homes and businesses. What is needed is open

(inter- )access to all essential facilities on all technologies at proper prices

and conditions.5

To access essential facilities, entrants or competitors pay an access price

(regulated) and the incumbent must grant access to all fi rms, including its

own divisions, on an equal non- discriminatory basis. Under this regime it

is important to get the access price right. If it is ‘too low’, then the incum-

bent has low and inadequate incentives to invest and innovate, a subsidy

is provided to the entrant, and ineffi cient entry is encouraged. If the access

price is ‘too high,’ competitors have an incentive to bypass ineffi ciently

the incumbent’s facilities, effi cient entrants are discouraged from entering,

and competition is not likely to prevail.

The promotion of effi ciency in network development investments is a

role and a responsibility that the telecommunications regulator should

assume. The main reasons are threefold:

Investments in network development are typically large, long- term ●

irreversible investments with signifi cant economies of scale, making

a given platform a non- standard natural monopoly.

The diff erent network development strategies off er signifi cantly dif- ●

ferent levels of embedded managerial fl exibility to delay or advance


the development timing (to slow down or speed up the pace of

development), to increase or decrease the planned network size and

scope, and to raise or reduce the planned quality of the network

infrastructure as new information on signifi cantly volatile demands

and costs is gathered over time.

The volatility of demand and costs together with the relative ●

embedded managerial fl exibility and the signifi cant irreversibility

of investments in network development require new evaluation

methodologies (real options valuation), which are still underused by

network operators.

In such a context, the regulator must make sure that inter- access to

essential facilities is available, at proper non- discriminatory conditions

and prices, to foster entry by more effi cient providers.6 The regulator must

also recognize that the proper non- discriminatory conditions and prices

must account for the real options that are exercised, as the network is

being built or developed.7 Eventually, the network developing fi rm is stuck

with the technological characteristics of its realized network. Exercising

those real options can represent a signifi cant cost in network develop-

ment and must therefore be accounted for in determining the proper non-

discriminatory access prices and conditions that will govern the use of the

incumbent’s network by competitors. Unless a proper account is made of

those real options- related costs, the development and maintenance of the

telecommunications networks are likely to be ineffi cient, thereby imposing

signifi cant real costs on society.

This leads to two important observations. First, in reviewing pricing

rules for inter- access to networks, the regulators must consider not only

the development of competing networks on alternative technological plat-

forms but also in each case the implicit costs of the real options exercised

in developing the network infrastructures. Second, in reviewing the rules

of competition between networks, the regulators should make explicit and

transparent the pro- competition special treatment that it wishes to grant

to some competitors, rather than directly impose price and technology

controls on others.

In a standard natural monopoly industry, profi ts are likely to rise

rapidly when the market expands since revenues then increase linearly or

proportionately with the number of subscribers while costs increase typi-

cally very slowly due to the presence of economies of scale, economies of

scope and network economies. The increase in profi ts means an increased

rate of return on capital, exceeding over time by a signifi cant margin

the opportunity cost of capital of the fi rm. In any normal industry, this

increased profi tability would favor the entry of additional fi rms into the


industry, but not necessarily in a natural monopoly industry. In this latter

type of industry, the potential entrants understand that what is important

and crucial for them is not the price they observe or the profi ts the incum-

bent fi rm is presently enjoying, but rather the price and profi ts that they

will be experiencing if and when they enter the market and compete with

the incumbent fi rm.

It is post- entry goods and services prices that preoccupy the potential

entrants, not actual pre- entry prices. In particular, if it is understood that

regulators are going to control pre- entry prices at a relatively low level but

that they will let prices be freely determined once new competitors have

entered the market, then observed pre- entry prices will have no eff ect on

the decision to enter or not. Potential entrants will evaluate what competi-

tive post- entry prices are likely to be and then decide if those prices justify

entry or not.

If the regulated pre- entry prices in the local wireline services market

are fi xed at a relatively low level in order to prevent monopoly pricing, it

may give the wireline incumbent a hedge insofar as these regulated prices

prevent the emergence of alternative bundles of services off ered over com-

peting technologies. If those technologies cannot profi tably compete with

the low prices set by the regulator for the local wireline services market,

then consumers will eventually suff er from the reduced technological

advances and the reduced variety in products and services. A short- term

benefi t from lower local wireline prices is obtained at the expense of a

long- term cost in terms of reduced choices, reduced innovation eff orts and

reduced competitive pressures.

If the regulated pre- entry prices in the local wireline services market are

fi xed at a relatively high level in order to prevent predatory pricing and

favour competition (through imposing fl oor prices on the ILEC without

regulating in any way the CLECs’ pricing, forbidding win- back actions by

the ILEC for a certain period, and other similar policies), it may give the

competitors (the CLECs) a hedge insofar as the regulated prices prevent

the incumbent from competing effi ciently. Such pricing regulation is likely

to prevent the emergence of a more effi cient portfolio of goods and serv-

ices off ered to customers, at least in the short run.8

We need a diff erent, more effi cient approach to controlling monopoly

pricing and predatory pricing while avoiding the subsidization of inef-

fi cient entrants, that is, a more effi cient approach to developing a level

playing fi eld in short- run competition while at the same time keeping a

high intensity of incentives for long- run competition. This more effi cient

approach rests on unregulated pricing of telecommunication services

subject to general pricing policies that the regulator can impose and

enforce on all players, ILECs and CLECs.


The current dominant approach to the prevention of monopoly pricing

(mainly in wholesale markets) and predatory pricing (mainly in retail serv-

ices) in networks rests on the notions of marginal, incremental or avoid-

able (MIA) cost. Although these three concepts of cost are not exactly the

same, they share similar foundations. The marginal cost is the additional

cost a fi rm must incur to increase its production of a given service by a

small amount; the incremental cost refers to the same concept for more

important increases; the avoidable cost refers to the reduction in cost a

fi rm would experience if it were to drop the production of a particular

good or service from its current portfolio of products and services (the

avoidable cost includes therefore the fi xed cost specifi c to that product or

service).

A fi rm is considered capable of exercising market power if its price for

a particular good or service could profi tably be set signifi cantly above its

MIA cost, including its MIA cost of capital, for a signifi cant period of

time. It could be considered guilty of predatory pricing if its price for a

particular good or service is below its MIA cost for that good or service,

including again its MIA cost of capital. Unfortunately, the use of the

concept of MIA cost is a source of major problems and confl icts. The

problems and confl icts stem from the fact that the MIA cost of a particular

good or service is a concept suff ering from fi ve major shortcomings when

economies of scale and scope as well as network economies are present,

as in telecommunications: it is non additive, order- dependent, unstable,

manipulable and horizon- dependent.

Clearly, a proper notion of cost allocation for the purpose of verify-

ing if a fi rm’s pricing behavior violates the non- predatory pricing rules

must satisfy the requirements of additivity, order- independence, stability,

non- manipulability and horizon- independence. One such cost allocation

procedure is the Shapley–Shubik (SS) cost allocation rule (Boyer et al.,

2006). In a nutshell, this rule corresponds to the average MIA cost over all

possible production orderings of the goods and services off ered by a fi rm.

Rather than taking simply the usual MIA cost, the SS procedure takes

the average MIA cost over all orderings, thereby making it immune to the

shortcomings of the cost allocation rule based on the MIA cost.

Moreover, the SS cost allocation rule has two other important proper-

ties, which prove very helpful in practice. First, it is such that the cost

shares are invariant to the decomposition of total cost between specifi c

and common or joint costs. This property is important because the decom-

position of the total costs between joint costs (to be shared) and specifi c

costs (to be borne by the specifi c good and service causing them) is very

often a source of confl ict. The fact that the decomposition is irrelevant is

therefore a property that is most welcome. Second, the SS cost- sharing


rule satisfi es demand monotonicity, which means that the cost allocation

of a good or service increases if the demand (production) of that good or

service increases while the production of others remain constant, and also

that under economies of scale the cost share of a good or service decreases

if the production of another good or service increases while its own pro-

duction remains constant. Again, this property is conducive to an agree-

ment between diff erent stakeholders. More generally, the cost allocations

obtained from applying the SS procedure can be shown to be an almost

costless surrogate for the cost allocation that would be obtained through

lengthy and costly bargaining ‘between the diff erent goods and services’

(or their demanders/consumers) as to the proper allocation of common

costs.

The regulator would consider a fi rm guilty of predatory pricing if the

fi rm was found to price one (group) of its products or services below the

per- unit cost allocation associated to that product or service in the total

cost of the fi rm. Hence, the per- unit cost shares become the price fl oors

under which a fi rm would not be allowed to price any of its goods and

services. Those price fl oors are determined by the ‘real’ cost function of the

fi rm itself and not by some argument regarding the necessity to grant some

advantage to competitors.

Combining proper access pricing for essential facilities and proper cost

allocation rules leads us to the intended result: if a fi rm is indeed more

effi cient than another, it should be able to achieve per- unit costs for its

goods and services that are in general lower than those of its competitors.

In that sense, the SS cost allocation rule is a pro- competitive price fl oor

rule, protecting all fi rms, incumbents and competitors against predatory

pricing. It creates adequate incentives for all fi rms to reduce the costs of

their respective goods and services, thereby developing a level playing fi eld

with similar competition rules for everyone. In the presence of economies

of scale, economies of scope and network economies, the SS cost alloca-

tion rule delivers the proper pricing benchmark compatible with the belief

that competitive pressures constitute the most important dynamic driver

toward socio- economic static and dynamic effi ciency. The experience

gathered over time in diverse applications of the SS cost allocation rule9

makes it a good candidate for the basic pricing rule. This does not mean

that the application of the procedure is (always) easy and straightforward,

but none of the alternative procedures, and certainly not the one presently

used, is easy and straightforward either.

How then can the government or regulator adopt a proactive strategy

to favour the emergence of more competition? The answer is simple: by

promising entrants a generous but deferred investment grant or tax credit

to be paid n years after entry conditional on the entrant’s acquisition of a


market share larger than x percent. Only entrants with superior products

or services, and/or superior technology, and/or better- quality consumer

service, or who believe that they can beat the incumbent one way or

another, will enter the industry.10

5.6 CONCLUSION

I have argued in this chapter that the traditional market shares measure

of competition is inadequate in the telecommunications industry because

this industry has more of the characteristics of an emerging industry than

of a mature industry. In such a context, the role of the regulator should be

focused on three responsibilities.

Firstly, acting as the trusted generator of information for the consumers

on pricing structures and product characteristics, underlying the derived

demand for telecommunications goods and services, namely connectivity,

fl exibility, safety, dependability, accessibility, capacity (high speed and

broadband) and user- friendliness.

Secondly, acting as the manager of the level playing fi eld conditions to

favor both static effi ciency and dynamic effi ciency, and enacting indirect

policies to control monopoly and predatory pricing, based on full cost

sharing between goods and services off ered by any given fi rm (as provided

for example by the cooperative game- theoretic SS cost allocation rule).

Thirdly, acting as the promoter of effi cient investment programs in

network development and maintenance to guarantee the integrity of the

global telecommunications network, and designing access pricing rules

incorporating all network access costs (including the real options costs

embedded in completed networks). Those redesigned responsibilities

of the regulator are in fact the end point of the sequence of regulatory

reforms that have been implemented since 1997. The telecommunications

industry regulator would become a truly pro- competitive watchdog and

an integral part of competition policy implementation.

Although the information gathering, the institutional reforms and

the change of minds and attitudes required to implement a true pro-

competitive regulatory framework for telecommunications, in particular

for the local access markets, represent a sizable and signifi cant undertak-

ing, it is no more and probably much less demanding than the prevailing

regulatory framework. Since the change in approach will take some time,

it is urgent for national authorities to launch a signifi cant program of

research, consultation and transformation, aimed at developing the neces-

sary means and procedures to implement eventually a switch to the new

pro- competition rules proposed in this chapter.


NOTES

1. This chapter is a revised and condensed version of my paper ‘The measure and regula-tion of telecommunications markets’ (CIRANO working paper 2005s- 35). I want to express my thanks to the ITS Montreal Conference participants and to the external referees for their comments.

2. The authors used household- level data from the Bill Harvesting dataset from TNS Telecoms ReQuest Market Monitor® along with its survey response (see www.tnstelecoms.com). The estimate for the cross- price elasticity of fi xed line telephony (all lines combined) from mobile telephony prices is lower (of the order of +0.06 to +0.08), as expected insofar as the main eff ect should be on second fi xed lines. See also Dzieciolowski and Galbraith (2004) and Zimmerman (2007). More recently, Taylor and Ware (2008, p. 6) estimated an all fi xed lines combined cross- price elasticity between wireline and wireless of 1.4.

3. The Economist (2003). The impacts of the emergence of Bertrand competitive condi-tions in an industry can be quite signifi cant: the Canadian- based Nortel saw its capi-talization fall from $400billion to $3billion in two fi nancial years 2000–2002, one of the main reasons being the large overcapacity in the industry. Nortel, North America’s biggest maker of telephone equipment, has fi led for bankruptcy protection under the US Chapter 11 bankruptcy protection law on 14 January 2009 (BBC News).

4. The 2005 Nobel Prize winners in economic science Thomas Schelling and Robert Aumann were pioneers in the development of such competitive modeling.

5. In proceedings under competition law, the defi nition and characterization of ‘essential facilities’ remain a hotly debated issue.

6. For a presentation and discussion of access pricing rules, see Boyer and Robert (1998), and Boyer et al. (2008).

7. For a discussion of the real options approach to evaluating investment decisions, see Boyer et al. (2004a). For a discussion of real options and access pricing in network industries, see Haussman and Myers (2002).

8. What is wrong with lower prices? Low prices, or more generally loss leaders, may be detrimental for competition if they are intended to harm competitors. If they are not, for instance if a telecommunications provider announces a sale of telephones at prices below costs in order to reduce costly inventories of slow- moving goods, then such low prices would be considered fair rather than predatory. In boxing, there is a diff erence between fi ghting and trying to beat the opponent by throwing hard punches to put him unconscious (legal, fair) and doing it by biting the opponent’s ear (illegal, unfair). In tennis, one is allowed to smash the ball as hard as one can, but not to aim at the oppo-nent’s head; similarly with fair low prices and predatory low prices.

9. See Young (1994) for discussions of some of those applications.10. For an application of such principles to spectrum auctions, see Boyer (2007).

REFERENCES

Boyer, M. (2007). Optimal policy relative to spectrum auction. Industry Canada, Government of Canada. Gazette Notice DGTP- 002- 07, May.

Boyer, M., M. Benitah and S. Weihao (2008). Real options, network development and network access. WIK NETCONOMICA 2008 Conference on Current Issues in Network Economics: Regulatory Risk, Cost of Capital and Investment Incentives. September, Königswinter, Germany.

Boyer, M., P. Christoff erson P. Lasserre and A. Pavlov (2004a). Value creation, risk management and real options. Icf Journal of Management Research, 3(10),


42–62; and in CIRANO 2003RB- 02 Burgundy Report at http://www.cirano.qc.ca/pdf/publication/2003EB- 02.pdf.

Boyer, M., P. Lasserre T. Mariotti and M. Moreaux (2004b). Preemption and rent dissipation under Bertrand Competition. International Journal of Industrial Organization, 22(3), 309–28.

Boyer, M., P. Lasserre and M. Moreaux (2007). The dynamics of industry invest-ments. CIRANO working paper 2007s- 09.

Boyer, M., M. Moreaux and M. Truchon (2006). Partage des coûts et tarifi cation des infrastructures. CIRANO Monograph 2006MO- 01.

Boyer, M. and J. Robert (1998). Competition and access in electricity markets: ECPR, global price cap and auctions. In G. Zaccour (ed.), Deregulation of Electric Utilities (pp. 47–74). Norwell, MA: Kluwer Academic Publishers.

Dzieciolowski, K., and J.G. Galbraith (2004). Indicators of wireline/wireless com-petition in the market for telecommunication services. CIRANO 2004RP- 21; Bell Canada, Analysis of Local Residential Voice Network Usage, a detailed study covering the two- month period 1 January to 29 February 2004.

The Economist, (2003). Beyond the bubble. 9 October.Haussman, J. and S. Myers (2002). Regulating the United States railroads: the

eff ects of sunk costs and asymmetric risk. Journal of Regulatory Economics, 22(3), 287–310.

Rodini, M., M.R. Ward and G.A. Woroch (2002). Going mobile: substitutability between fi xed and mobile access. Center for Research on Telecommunications Policy Working Paper CRTP- 58, Haas School of Business, University of Berkeley. December.

Taylor, W.E. and H. Ware (2008). The eff ectiveness of mobile wireless services as a competitive constraint on landline pricing: was the DOJ wrong? NERA Working Paper, December 2008.

Young, H.P. (1994). Cost allocation. In R.J. Aumann and S. Hart (eds), Handbook of Game Theory Vol. II (pp. 1191–1235). Amsterdam: North- Holland.

Zimmerman, P. (2007). Recent developments in US wireline telecommunications. Telecommunications Policy, 31(6–7), 419–37.

128

6. Preventing harm in telecommunications regulation: a new matrix of principles and rules within the ex ante versus ex post debate1

Kenneth Jull and Stephen Schmidt

6.1 INTRODUCTION

A recent tragedy involving the 911 emergency service illustrates the need

for regulation in those sectors that impact human health and safety. A

family who subscribed to Voice over Internet Protocol service moved

from Toronto to Calgary. When their young child became gravely ill, they

called 911 and waited for an ambulance that did not arrive. Tragically,

the 911 operators dispatched an ambulance to the Toronto address, as

it had not been updated to refl ect the move to Calgary. Sadly, the young

child passed away. It is not our intention to comment on the specifi cs of

this case. We can comment, however, that this case underlines the need

for regulation in those sectors where the potential for human harm is the

greatest. It is of little comfort to a family who has lost a child to advise that

the situation will be remedied in an ex post world, if it might have been

prevented by ex ante measures to ensure comprehensive and accurate 911

coverage across all telecommunications networks. In a six- month investi-

gation of Canada’s 911 service, the Globe and Mail found that: ‘a lack of

federal oversight, regulatory loopholes and outdated technology have left

this country’s emergency dispatchers scrambling to locate callers who dial

911 from cell phones or from Internet phones’.2

Most North American regulatory systems, in areas ranging from the

environment to securities law, utilize a basic ex post model: the regula-

tion or statute sets the standard, and if one breaches the standard, then

remedies are imposed after a legal fi nding of wrongdoing. By contrast,

ex ante systems operate prospectively, depend on prior approval and

impose remedies on an anticipatory basis, prior to any actual breach or

Preventing harm in telecommunications regulation 129

wrongdoing. Regulatory systems in the telecommunications industry have

traditionally been ex ante systems. For example, a review of Canada’s

Telecommunications Act shows a regulatory scheme that is weighted

towards ex ante regulation through, among other things, tariff approval

requirements, agreement approval requirements, unbundling orders,

interconnection orders, and so on. The basic presumption of the Act is that

all telecommunications services are governed by tariff s that must be fi led

and approved by the Canadian Radio- television and Telecommunications

Commission (the CRTC) prior to the off ering of service, unless the CRTC

has made a decision expressly to forbear from exercising its powers in

certain circumstances (Government of Canada, 2006). In other words,

telecommunications services cannot be off ered to the public without prior

approval by the state.3

The ex ante system of prior approval and supervision by the regulator

has been criticized by commentators as being a blunt ‘one size fi ts all’

mechanism (Janish, 2005, p. 13). A new regulatory framework has been

recommended by the Telecommunications Policy Review Panel (TPRP)

Final Report, released in March 2006 (Tremblay et al. (2006). The TPRP

approach would set out broad principles to prohibit anticompetitive

conduct instead of detailed ex ante rules.

We propose three principles that ought to govern the balance between

ex ante and ex post systems in the telecommunications world of the

future. First, governments and regulators ought to focus on the preven-

tion of harm and the serving of human needs in the order of their priority

(Archibald, Jull and Roach, 2004). Recent scholarship has focused on the

prevention of harm in regulation. In his book The Character of Harms,

Malcolm Sparrow (2008) argues that regulators should focus on fi xing

the big problems that cause the most harm, by developing strategies that

operate at the operational risk level (Sparrow, 2008). We see a role for ex

ante regulation in preventing harm in the telecommunications sector (such

as preserving the integrity of emergency 911 systems) and in protecting

vulnerable groups within the meaning of human rights legislation. The

application of harm theory to telecommunications policy puts the industry

into the proper context when considering the role of regulation. In short,

telecommunications may prevent harm from coming to people, but not

to the extent of some other sectors (such as health, the environment and

transport) that may warrant deeper regulatory oversight.

The CRTC has responded to the 911 issue by announcing that it will force

the cellphone industry to upgrade Canada’s 911 systems. Government offi -

cials said in January 2009 that they will impose a February 2010 deadline

to install the necessary equipment to give 911 dispatchers the ability to

locate cellular calls in an emergency.


In announcing the decision to the media, the CRTC uses language which

is consistent with the thesis of this chapter: ‘I thought that this would put

the matter to rest’, a CRTC offi cial said in an interview. ‘We are concerned

about the safety and security of Canadians . . . that’s sort of our guiding

light if you wish’ (Robertson, 2009).

The second principle that we propose is that regulation of competitive

markets will encompass risk management and social regulation.4 The

fi nancial crisis of 2008 underscores the need for regulation of risk manage-

ment systems in economic markets.

Our third principle is that the regulatory state should off er multiple

models that refl ect the diff erent needs and interests at stake within both

the ex ante and the ex post world. Indeed, there is a growing recogni-

tion in legal circles that any sophisticated system ought to off er a full

range of models based on the seriousness and economics of the case

(Mullan, 2005). For example, within the ex post paradigm, there is a

new principles- based model emerging, led by developments in the securi-

ties fi eld. Ford (2008) describes this development and she identifi es the

principles- based and outcome- oriented regulation that has been advo-

cated by the British Columbia Securities Commission as an example

of the new governance. She describes the basic diff erence in the two

systems as follows:

The classic example of the diff erence between rules and principles or standards (to use another term) involves speed limits: a rule will say, ‘Do not drive faster than 55 mph’, whereas a principle will say, ‘Do not drive faster than is reason-able and prudent in all the circumstances’. Put another way, a rule generally entails an advance determination of what conduct is permissible, leaving only factual issues to be determined by the frontline regulator or decision- maker. A principle may entail leaving both specifi cation of what conduct is permissible and factual issues for the frontline regulator. (Ford, 2008, p. 8)

If one accepts that the ex post model might have two subsets within it,

being rules- based and principles- based, then one must also accept that

the ex ante paradigm also has two subsets within it, being rules- based

and principles- based. The dichotomy of ex ante and ex post becomes

a matrix of four models. An example of a rules- based ex ante solution

in the highway traffi c example would be the requirement that all trucks

have governors installed that prevent them from travelling over 105 km

per hour. An example of a principles- based ex ante solution would be the

requirement that each driver pass a driver’s test that is based upon criteria

developed from a principled approach.

Table 6.1 illustrates the integration of the three principles that we are

advocating.


6.2 PRINCIPLES OF REGULATION

6.2.1 First Principle: Governments and Regulators Ought to Focus on the

Prevention of Harm and the Serving of Human Needs in the Order

of their Priority

Types of harm

Sparrow (2008) argues that regulators should focus on fi xing the big prob-

lems that cause the most harm, by developing strategies that operate at the

operational risk level. He further observes (p. 10) that the fi nancial sector

laid an early claim to the phrase ‘risk management’. Risk management is

the process of weighing policy alternatives in light of the results of risk

assessment, and selecting appropriate control options, including regula-

tory measures (Powell, 2000, p. 138). Sparrow’s (2008, p. 11) major con-

tribution is to inject the word ‘harm’ into the risk formula, which includes

potential harms. He identifi es fi ve properties of harm that are common to

many disciplines: invisible harms, harms involving conscious opponents,

catastrophic harms, harms in equilibrium and performance- enhancing

risks (p. 173). Each is worth briefl y reviewing.

Invisible harms may be invisible by design, such as white- collar crimes.

In other cases the harms are not visible because of a lack of reporting

will or mechanisms. Sparrow (2008, Chapter 8) gives the example of the

internal revenue service failing to measure the problems of non- fi lers, as by

defi nition there was no record of those who did not fi le their tax returns.

Table 6.1 Ex ante and ex post rules- and principles- based regulation

matrix

Paradigm

Timing

Ex Ante Ex Post

Rules- based Harm related: imposition of

detailed technical rules for

operation of 911 emergency

services as part of licensing

requirements.

Prohibition of contacting

customers on a do- not- call

list.

Principles- based Imposition of licensing terms

requiring reasonable, non-

discriminatory access to

services for disabled

persons.

Principle that carriers must

interconnect with each

other for the exchange of

traffi c on reasonable terms.


One solution suggested by Sparrow would be to cross- reference outside

databases. For example, a review of the Yellow Pages of businesses in a

particular area could be cross- referenced to the number of businesses who

actually fi led.

In the telecommunications fi eld, an example of invisible harm might be

the exclusion of certain disadvantaged groups from service due to cost or

geographic issues. As there is no record of non- subscribers, the harm to

them is not easily measured but can only be speculated on.

Conscious opponents are described by Sparrow (2008, Chapter 9) as

having a brain behind the harm. The opponents can range from fraud per-

petrators to students cheating in exams. Sparrow observes that these types

of harms confound our normal reliance on probabilities, as it is diffi cult to

predict the intentional conduct of opponents. The methods of fi ghting this

type of harm are traditionally found in the criminal justice system: use of

surveillance, informants and confessions.

In the telecommunications fi eld, as it has been historically highly regu-

lated, it is likely that the incidence of harm caused by fraudulent behavior

of service providers is relatively low. As the market becomes increasingly

deregulated, this could potentially change. Although this is speculative,

we predict that the level of intentional or fraudulent conduct will continue

to stay relatively low as the market is deregulated, because the culture

of most players in the industry has its origins in regulated compliance.

This does not mean however that regulators should be complacent as the

market becomes deregulated. Regulators must attempt to measure the

possibility of non- compliance and be prepared to borrow techniques from

the criminal justice area.

Catastrophic harms are rare events with very serious consequences,

such as nuclear terrorism or earthquakes. In such events the telecommu-

nications networks may be threatened and must have in place contingency

plans and backup systems.

Harms in equilibrium relate to the imbalance that may be created in a

competitive industry from varying levels of compliance. Sparrow (2008,

Chapter 11) gives the example of an industrial sector largely out of com-

pliance with environmental regulations. A company that desires to be

compliant will be put at a disadvantage from a cost perspective relative to

its peers who are non- compliant.

Performance- enhancing risks relate to the organization’s performance

goals. Sparrow (2008, Chapter 12) gives the example of aggressive, fraudu-

lent or abusive billing practices which enable a corporation to maximize

profi ts. Sparrow argues that departments responsible for compliance

within an organization must have credibility if the organization is to resist

performance- enhancing risks. The reduction of performance- enhancing


risks relates directly to compensation. Compensation ought to be (and

historically has not been) related to levels of compliance in addition to

levels of economic gain.

Ranking of types of harm

One of the authors of this chapter has argued elsewhere that where leg-

islatures codify risk assessment values that potentially impact on human

life and health, the legislative branch should utilize the wealth of scientifi c

expertise to the greatest extent possible (Archibald, Jull and Roach, 2004).

The uneasy tension between politics, the public and experts’ assessments

has led to several problems in risk management (Slovic, 1999). A problem

that emerges from this tension is identifi ed by Justice Breyer (1993) as

‘random agenda selection’:

Agency priorities and agendas may more closely refl ect public rankings, politics, history, or even chance than the kind of priority list that environmental experts would deliberately create. To a degree, that is inevitable. But one cannot fi nd any detailed federal governmental list that prioritizes health or safety risk prob-lems so as to create a rational, overall agenda, an agenda that would seek to maximize attainable safety or to minimize health- related harms.

Breyer (1993) illustrates a very interesting comparison of the expert’s

assessment of various risks, against the public’s assessment of risks, in

Table 6.2.

The public policy implications of the above comparison are enormous.5

The diff erences between expert and public assessment of risk are the

subject of analysis by the Canadian Centre for Management Development

(Hill, 1969). Serious issues such as air pollution, which the experts assert

pose high risks to public health, are not viewed by the public as signifi cant.

People rank vehicle exhaust as fi fteenth in priority, and yet the experts

indicate that this problem should be ranked much higher.

Obviously, there is a serious knowledge gap between the public percep-

tion of risk and the experts’ perspective. In setting regulatory standards,

we invite the legislative branch to measure the potential gravity of impact

upon our society in accordance with scientifi c methodology. In other

words, governments should try to narrow the information gap between

the experts and the public, by taking the lead in recognizing new scientifi c

theories. Sparrow (2008, p. 13) makes a similar argument that: ‘At this,

the highest level, where one adopts the more paternalistic perspective of

governments, the psychology of individual risk perception matters less, or

matters for diff erent reasons.’ Science can identify a hierarchy of human

needs within social psychology. Governments and regulators should be

measured by their ability to serve these needs in the appropriate order.6


At this juncture, the reader may say, what does all of this have to do

with telecommunications law? Several conclusions relevant to the prac-

tice of telecommunications regulation fl ow from the foregoing. First,

telecommunications regulation ought to have as its primary focus the

prevention of harm. The approach advocated by Sparrow (2008, p. 17)

of preventing bad things as opposed to the construction of good things

brings perspective to the progressive lessening of sector-specifi c economic

regulation in telecommunications. Simply put, we should ask whether

Table 6.2 Knowledge gap between the public perception of risk and the

experts’ perspective

Public EPA experts

1. Hazardous waste sites

2. Exposure to worksite chemicals

3. Industrial pollution of waterways

4. Nuclear accident radiation

5. Radioactive waste

6. Chemical leaks from underground storage tanks

7. Pesticides

8. Pollution from industrial accidents

9. Water pollution from farm runoff

10. Tap water contamination

11. Industrial air pollution

12. Ozone layer destruction

13. Coastal water contamination

14. Sewage plant water pollution

15. Vehicle exhaust

16. Oil spills

17. Acid rain

18. Water pollution from urban runoff

19. Damaged wetlands

20. Genetic alteration

21. Non hazardous waste sites

22. Greenhouse eff ect

23. Indoor air pollution

24. X- ray radiation

25. Indoor radon

26. Microwave oven radiation

Medium- to- low

High

Low

Not ranked

Not ranked

Medium- to- low

High

Medium- to- low

Medium

High

High

High

Low

Medium- to- low

High

Medium- to- low

High

Medium

Low

Low

Medium- to- low

Low

High

Not ranked

High

Not ranked

Source: Breyer (1993), Table 4, p. 20, as reproduced with permission in Archibald, Jull and Roach (2004).


it is wise to put signifi cant public resources into the detailed economic

regulation of telecommunications when that money might be spent more

wisely in other sectors that prevent harm to people. These sectors include

health, the environment and security. (Of course, where issues of harm

genuinely present themselves in telecommunications markets, they do

merit regulatory attention.) Second, the latest scientifi c thinking can, and

should, be invoked to focus harm- prevention eff orts in accordance with

human needs priorities. This means, for example, that a regulatory focus

on matters such as the 911 service which have a dramatic impact on saving

lives will have signifi cant priority over measures focused, for example, on

the pricing of communications services (where questions of human harm

are less acutely engaged). Third, telecommunications regulatory practice

organized around the prevention of harm, and duly informed by scientifi c

thinking, will look fundamentally diff erent from the practice of regulation

today where public and private investments in regulation almost certainly

refl ect an inversion of this thinking.

We acknowledge that the pricing of telecommunications services can,

in and of itself, engage human harm issues. For example, this would be

the case if services determined to be essential to participation in society

and the economy were priced at a level, or in a manner, that precluded

customers from engaging in social and economic activity via the telecom-

munications network. There is a general recognition that local telephone

service is an essential service and that it ought to be priced in a manner

that makes it accessible to all users, including lower- income users. That

being said, market circumstances in Canada have evolved to the point

where the CRTC relies primarily on competitive markets to constrain the

pricing of local telephone service. In sum, we acknowledge that supplier

pricing behavior for the subset of telecommunications services deter-

mined to be essential services can raise human harm issues. This is why

essential services have been historically subject to economic regulation

where markets forces, alone, were not considered suffi cient to constrain

pricing.

6.2.2 Second Principle: Regulation of Competitive Markets will

Encompass Risk Management and Social Regulation

The second principle that we propose is that the regulation of competitive

markets will encompass risk management and social regulation. The 2008

fi nancial crisis emphasized the need for regulation of risk management

systems in competitive markets. A leading text used in business schools to

teach risk management is Seeing Tomorrow by Ron Dembo and Andrew

Freeman (1998). Ten years before the 2008 crisis, Dembo and Freeman


(p. 28) warned that: ‘entire businesses have been brought to their knees by

single- scenario forecasting’.

Regulation cannot of course impose prudent investment strategies, but

what it can do is require that a risk assessment structure is in place and

that investors have maximum knowledge of the risks that they may face.

The US Securities and Exchange Commission has recently encouraged

companies to take a more risk- based approach to complying with legisla-

tion such as the Sarbanes–Oxley corporate reform package (Globe and

Mail, 2007). Following the fi nancial crisis of 2008 there is a need for clari-

fi cation about proper risk management techniques versus risk assessment

which merely measures the level of risk. Risk management is the process

of weighing policy alternatives in light of the results of risk assessment,

and selecting appropriate control options, including regulatory measures.

Risk management identifi es the priorities of an organization and in par-

ticular when audits of these priorities may be required (Archibald, Jull and

Roach, 2004, Chapter 7).

Applying the second principle to the reform of telecommunications

requires regulation of risk management in competitive markets. The

tariff scheme in the present Telecommunications Act requires prior

approval, which increases the risk of discouraging legitimate competi-

tion. The TPRP Final Report (Tremblay et al., 2006) cites Posner (1999)

for the proposition that economic regulation leads to activity to circum-

vent those constraints, which is a waste from the point of view of the

economy.7

Increased reliance on market forces brings the telecommunications

sector in line with other competitive industries. The TPRP recognizes

that new principles and rules must be drafted that will then be enforced:

‘Consistent with the deregulatory approach adopted by the Panel, the new

regulatory framework should set out broad principles to prohibit anti-

competitive conduct instead of detailed ex ante rules’ (Tremblay et al.,

2006, p. 22).

Returning to the TPRP recommendation, what will the broad principles

that will prohibit anticompetitive conduct look like? The starting point is

the existing Competition Act:

To address anticompetitive conduct in the telecommunications market, the provisions of the current Competition Act cannot be adopted word for word but can serve as a framework. Telecommunications is a network industry, with large sunk costs and signifi cant economies of density and scope as well as positive externalities. In such an industry, network eff ects are important, which naturally allow some players to have very large market shares in equilibrium. As well, the defi nition of the proper market for further analysis can be particu-larly diffi cult in telecommunications markets. (Tremblay et al., 2006, p. 24)


The lessening or removal of sector- specifi c economic regulation does

not require the abandonment of social regulation. Historically, globaliza-

tion forced deregulation in some industrial sectors, but at the same time,

expanded regulation occurred in ‘framework sectors’ such as the environ-

ment. The Canadian government has recognized this distinction with its

direction that the CRTC, ‘when relying on regulation, use measures that

are effi cient and proportionate to their purpose and that interfere with the

operation of competitive market forces to the minimum extent necessary

to meet the policy objectives’ (Governor in Council, 2006).

The TPRP recommended a new approach to control anticompetitive

conduct in telecommunications markets on the basis of complaints made

on an ex post basis, rather than by prescribing detailed ex ante restrictions

governing the provision of services. We argue in the next section that a

two- model dichotomy is not suffi cient, but rather that there is a matrix of

four models to choose from.

6.2.3 Third Principle: The Regulatory State should Off er Multiple

Models that Refl ect the Diff erent Needs and Interests at Stake

within both the Ex Ante and the Ex Post World

Defi nition of terms: ex ante versus ex post

There is some confusion about the distinction between ex ante and ex post

systems. We wish to defi ne our terms. The central concept that underlies

an ex ante system is prior approval of an activity. For example, the CRTC

may require that a tariff be approved, or a license be obtained, prior to the

provision of telecommunications services to the public. Rates, terms and

conditions for the provision of service are imposed ex ante, but any breach

of these may be addressed ex post. An ex post approach is distinguished

from an ex ante approach by the fact that the former imposes remedies

only after a legal fi nding of wrongdoing, whereas the latter imposes rem-

edies, preventatively, in advance of any wrongdoing.

These concepts may seem foreign, as the practical reality is that there

are very few telecommunications- related enforcement actions brought ex

post in Canada, despite the fact that there are a range of remedies pres-

ently available to the CRTC. We suspect that the same situation applies

globally. As a result of low enforcement, the perception may be that the

existing framework of prior approval is toothless. This paucity of enforce-

ment activity may simply refl ect high levels of compliance.

Other areas of the law also use a combination of ex ante prior approval

by a regulator combined with ex post enforcement. For example, in the

area of environmental conservation, it is a requirement of Canadian land-

owners that they seek permits for any work adjacent to fl oodplain areas.


Unlike telecommunications, the incidence of non- compliance is higher.

Regrettably, some landowners choose to ignore or violate the permit

process and undertake development without permits. In such a case, the

landowner can be prosecuted for committing the regulatory (strict liabil-

ity) off ence of development without a permit. In this area, it makes sense

that a specialized regulator may give ex ante approval for development in

sensitive areas, and that a failure to seek such approval will be enforced

with a regulatory off ence that deals with ex post conduct.

In the area of telecommunications, most are familiar with the type of ex

ante orders made through various mechanisms including tariff approval

and licensing processes. Less familiar are the ex post remedies presently

available to enforce regulatory non- compliance. There are remedies pro-

vided by the Telecommunications Act that may be enforced by courts,

and these broadly include regulatory off ence proceedings and damages

awards, based upon a statutory breach of the Act.

The TPRP recommended a new approach to control anticompetitive

conduct in telecommunications markets on the basis of complaints made

on an ex post basis, rather than by prescribing detailed ex ante restrictions

governing the provision of services. By defi nition, ex post enforcement

occurs only subsequent to a violation of rules, and is often triggered by

a complaint. The ex post model is used in most legal systems, ranging

from regulatory to criminal law. A pure ex post system has its origins in

the common law system of privately enforced rights. Here, the focus is on

redressing the harm caused by a breach of rules or standards.

Where health and safety of consumers is at risk, we can envision some

justifi ed need for ex ante approval by a regulator

Ex ante control has the advantage of using a rational approach towards

an issue in advance (Yael Aridor Bar- Ilan, 2007). The CRTC has rec-

ognized this by recently reissuing its 911 requirements for Voice over

Internet Protocol (VoIP) service providers. In its 911 Circular, the CRTC

articulates detailed rules that must be followed by VoIP providers when

handling 911 calls. This would be an example of a rules- based ex ante

system, as it would be a requirement for all telecom providers. In the case

of the 911 service, ex ante rules concerning the service could be coupled

with additional measures to ensure compliance (periodic audits or inspec-

tions, for example).

There is precedent for the use of ex ante prior approval to protect

human life. For example, in Canada one must obtain prior approval from

a regulator to build in areas adjacent to fl oodplains, given the risks to

human life and safety that may be created by structures that are in the path

of a fl ood. As a general principle, our reference to health and safety is not


restricted to life- threatening situations. We use the term in the tradition

of prevention of harm to human health, which is a wide- ranging concept

and is consistent with Sparrow’s concepts. In particular, prevention of

catastrophic harms justifi es an ex ante approach.

Where vulnerable minority groups8 within the meaning of human rights

legislation may not have appropriate access to telecommunications, an ex

ante approach fi lls the gap

Disabled persons, for example, may not have the resources to enforce their

rights in an ex post world, depending upon the level of regulatory enforce-

ment. Moreover, access to telecommunications for disabled persons may

present special needs that are more appropriately dealt with ex ante.

Sparrow might characterize the risk to disabled people as an invisible

harm, as they are less likely to have the resources to lobby for coverage.

The CRTC Discussion Paper (Government of Canada, 2006) submitted

to the Telecommunications Policy Review (TPR) noted that competition

law does not contain mechanisms to ensure access to telecommunications

networks by disabled persons. We recognize a limited role for ex ante regu-

lation to protect groups such as the disabled, in conjunction with ex post

remedies (discussed later in this chapter).

The choice between rules- based and principles- based systems will be

assisted by analyzing the amount of fl exibility that is required

Within both the ex ante and ex post paradigm, there may be both rules-

based and principles- based models. We believe that this distinction

may have been lost in the fray of the fi ght between ex ante and ex post

approaches. Every regulatory framework can be placed along a continuum

between rules- based and principles- based. Cunningham (2006) made this

observation in his study of Principles and rules in public and professional

securities law enforcement: a comparative US–Canada inquiry:

Rules and principles are both necessary and desirable features of an overall securities regulation system. Whether laws are better cast in one form or the other depends on trade- off s involving certainty versus contextual judgment on the one hand and the relative novelty versus norm recognition on the other. Aspirations to create securities regulation regimes denominated as principles- based or rules- based likely are conceptual fantasies, given the nature, complex-ity and capaciousness of these labels.

As noted above, a rules- based ex ante model would be the require-

ment that all trucks must have governors on them that prevent the truck

speed from exceeding 105 kph (Ministry of Transportation, 2008). Such a

rule is proactive, and protects the lives of drivers on the road. The rule is


overinclusive, in that it includes truck drivers who have impeccable driving

records and who are safe drivers. The rule is necessary, however, because

recent tragedies have demonstrated that there are some truck drivers who,

for various reasons, are not obeying speed laws. Given the size of the

trailers carried by trucks, the results of excessive speed can be deadly. The

requirement of governors on all trucks obviously lacks fl exibility. By con-

trast, a principles- based license is granted by a regulator (or examiner in

the case of transport) executing discretion based on principles of safety.

Sparrow (2008, Chapter 6) devotes a chapter of his book to the puzzles

of measurement. This novel analysis of measurement of harm contains

several insights for the ex post versus ex ante debate. In an ex ante world,

the focus is on measurement of process, being the process of regulatory

approval. In an ex post world, the focus shifts to measurement of harm.

This raises an important issue about resources. A shift from ex ante regu-

latory approvals will free up resources to be devoted to the prevention of

harm in an ex post world. It is essential that the resources be devoted

appropriately to ensure the robust nature of the ex post system.

In telecommunications, as noted above, there is a strong case for ex

ante regulation to ensure eff ective access to telecommunications services

by disabled persons. In an area of rapidly evolving technology – cellular

and PDA (personal digital assistant) handset design, for example – it may

not make sense to adopt an exclusively rules- based, ex ante approach

because overly prescriptive rules about device design9 will rapidly fall out

of date as technology evolves. In this case, a principles- based approach

may be preferable. This approach would impose principles- based acces-

sibility requirements on an ex ante basis. An example of this would be

an obligation that communications devices become reasonably accessible

to various classes of disabled persons. This principles- based approach to

accessibility refl ects the approach of Canadian human rights legislation.

Where a human needs analysis indicates a requirement for a more precise,

consistent and industry- wide implementation of an accessibility require-

ment (as might be the case with the Braille inscriptions on telephone keys),

a rules- based, ex ante approach would be suitable. This approach could,

of course, coexist alongside principles- based ex ante requirements for

accessibility.

A rules- based system, whether it is ex ante or ex post, is better suited to

regulated actions that are simple, stable and do not involve huge economic

interests.10 Rules may suff er from being either overinclusive or underinclu-

sive, due to the static nature of the legislative process. Economists predict

that one will fi nd more overinclusive rules in areas where the sanction is

simple damages or where transaction costs are low (Ehrlich and Posner,

1974).


Industries like telecommunications are undergoing rapid and funda-

mental technological and business transformation with the result that

they are not simple or stable and the economic stakes are very high. In

such contexts, a principles- based approach may be more appropriate.

Cass Sunstein (1995) argues that rules can be overtaken by changing cir-

cumstances in fi elds such as telecommunications: ‘In the face of rapidly

changing technology, current rules for regulation of telecommunications

will become ill- suited to future markets. For this reason it may be best to

avoid rules altogether or at least create only a few simple rules that allow

room for private adaptation’ (Sustein, 1995, pp. 993–4).

As is clear from our taxonomy, we would not go so far as rejecting all

rules in the telecommunications area. Braithwaite’s (2002, p. 54) study of

rules versus principles makes this sector- specifi c observation: ‘What a tel-

ephone means today may be something quite diff erent tomorrow. In terms

of the conclusion of this article, so long as the principles that underpin the

redefi ning are clear, redefi nable rules can regulate a transitional technol-

ogy with more certainty than fi xed rules’.

Some have argued that principles and standards are more conducive to

the development of an altruistic state (Goebel, 1993, p. 68). The debate

about our ability to control the world ex ante is a debate that has con-

tinued throughout human history. Aristotle, in his Nicomachean Ethics,

recognized that all universal laws made up in advance refl ect human

capacities and accordingly are going to be defective in addressing particu-

lar decisions (Bar- Ilan, 2007, p. 936). For example, where economic regu-

lation is concerned, it is extremely diffi cult, if not impossible, to predict

market outcomes before the fact.

In summary, there is a continuum from ex ante to ex post, with both

models using a combination of rules and principles where appropriate.

Where economic regulation is concerned, it is diffi cult, if not impossible,

to predict market outcomes before the fact. In those sectors where rules

are regulating economic decision making, a consistent application of a

prescriptive ex ante system will have major disadvantages. Instead, it is

likely that a menu of regulatory models will be more appropriate and that

the choice of model will vary according to the level of competition in the

relevant market

A more fl exible regulatory approach is justifi ed in recognition of the

dynamic and complex nature of the market, the waning of market power,

and the ability of entrants to act as a check on the behavior of the incum-

bent. An ex post regime maximizes fl exibility without abandoning regula-

tory control.

A rules- based, ex ante approach will be appropriate in the most extreme


cases of market failure, for example, in those rare instances where a carrier

has control over an essential facility. In such an instance, establishment

of non- discriminatory access to that facility will require the imposition of

ex ante rules governing the rates, terms and conditions. At the other end

of the regulatory spectrum are classes of services that are competitive, if

not forborne.11 For these services, greater reliance can be placed on the

market as a disciplining force and a more fl exible regulatory approach will

be justifi ed. Indeed, the dynamic nature of these markets strongly counsels

against a heavy- handed rules- based, ex ante approach because the costs

imposed by regulation in such circumstances outweigh the benefi ts to cus-

tomers. In this latter case, a principles- based requirement that rates be just

and reasonable in the circumstances, coupled with ex post enforcement,

may be suffi cient to protect the interests of customers.

A major theme of this chapter is that the ex post paradigm is not mono-

lithic, but in fact contains a spectrum of diff erent models that range from

rules- based to principles- based. There is a tendency in some quarters to

view the ex post world as a version of the old Wild West, where lawlessness

prevailed. The opposite is true, as the ex post paradigm governs criminal

and regulatory law which has historically protected our society. Ex post

systems are also preventative in nature and do not only apply after damage

is done. The motorist who drives by a person who has been stopped by

the police for speeding will be subject to general deterrence and may slow

down accordingly. This is an example of the proactive power of deterrence

within an ex post model. Sparrow (2008, Figure 6.1, p. 137) makes this

point by developing the concept of the unfolding chronology of harm.

Diff erent types of preventative steps may be separated by short periods of

time in this chronology and may be drawn from both ex ante requirements

or ex post rules and principles.

In those areas of the economy where technology is moving very quickly, an

ex post, principles- based system is likely superior to an ex post rules- based

system, as it maximizes fl exibility

Ex post rules, as distinct from principles, may suff er from infl exibility,

related to the slow pace of legislative reform. Accordingly, a principles-

based system may be superior. Interconnection in a multi- network, next

generation environment provides a good example of this. The fast- changing

and heterogeneous nature of the environment, with many providers, many

platforms (wireless, cable, copper) and many diff erent technologies within

those individual platforms, demonstrates the limits of both ex ante and

rules- based ex- post systems. Neither can fully anticipate, and adequately

codify, the range of interconnection arrangements that are likely to emerge.

In this type of situation, it makes more sense for a specialized tribunal to


articulate a principle (carriers must interconnect with each other for the

exchange of traffi c on reasonable terms). This specialized tribunal would

then have an ex post discretion in the interpretation of reasonable stand-

ards (relative to the implementation of the principle) after a carrier had

made out a case that the principle was being breached.12

A rules- based ex post system is suited to certain types of ‘social’ regulation

as contrasted to ‘economic’ regulation

The recent legislation relating to a ‘do not call list’ refl ects the ex post

rules- based approach. If a telecommunications provider fails to comply

with a direction to not call, this may be the result of inadvertence or it

may be the result of conscious harm, to use Sparrow’s term. Moreover,

unless all telecommunications providers are treated equally, there is a risk

of equilibrium harm.

A rules- based system may be very precise, depending upon how the

statute is drafted. An archetype of precise standards is found in the occu-

pational health and safety legislation. For example, the regulations require

that a guardrail system ‘shall be located at least 0.9 meters but not more

than 1.1 meters above the surface on which the system is installed’.13 The

occupational health and safety regulations are not so much prohibitions

as detailed prescriptions as to how one ought to conduct business. These

rules are an example of the potential for rules to educate, as well as making

prosecutions easier to win (Baldwin, 1990).

There are some disadvantages to precision. Rules may come close to an

ex ante model of control, if they start to resemble detailed licenses. One

could argue that occupational worker protection law should be ex ante,

and the only reason that it is not is a historical anachronism. Conversely,

telecommunications has largely been ex ante also as a result of histori-

cal antecedents. Back to the point of the chapter, we argue that history

ought not to unreasonably guide or bind us, but rather help us choose

the best model. Perhaps worker safety regulations ought to consider an

ex ante model requiring the fi ling of business proposals before construc-

tion, which set out a plan of safety, given the seriousness of the issues at

stake. In some cases, the technological change may result in standards

being set too low (an example would be an environmental scenario where

better methods are available to remove or prevent contamination). In

other cases, technological change may make the regulations redundant, or

worse, counterproductive.

In summary, as applied to telecommunications, there is a fear in some

quarters that deregulation in favour of an ex post system will lead to the

equivalent of a lawless Wild West. Precise rules, such as occupational reg-

ulations, demonstrate that this need not be the case. In the areas of social


regulation for telecommunications, such as health and safety regulations

for wireless radio equipment (transmitters, receivers, handsets), there is a

justifi ed basis for detailed ex ante rules coupled with ex post monitoring

and enforcement.

The type of rule, as explored above, should not be confused with the

appropriate remedy for the breach of the rule

There is a danger in the ex ante versus ex post debate of simplifying the

remedy issues. For example, some have argued that a disadvantage of the

ex post framework is that it is slow to react to damage that has occurred,

which may cause irreparable harm. This criticism fails to consider injunc-

tive relief that is available in the ex post model to prevent irreparable

harm. Quasi- judicial orders are exemplifi ed by the existing power of the

CRTC to make mandatory and restraining orders under section 51 of the

Act. In an ex post world, injunctions are a well- known remedy which have

developed their own body of precedent (Sharpe, 1998).

Telecommunications regulators will have a menu of potential remedies

available to them, including administrative monetary penalties, regulatory

fi nes, and criminal sanctions. The full spectrum of remedies should be

considered and suitably applied to address actual or potential breaches of

regulations. This leads to our last point.

As a general approach to remedies within the ex post world, we advocate an

approach towards enforcement that has its origins in restorative justice

Braithwaite (1985), a pioneer in the fi eld of restorative justice, devel-

oped the concept of a regulatory pyramid in his book published in 1985.

Ayres and Braithwaite (1992) elaborated further on the enforcement

pyramid. The base of the pyramid was concerned with persuasion of

the regulated by the regulators. If this approach proved ineff ective, the

regulatory eff ect could be incrementally escalated through warnings, civil

penalty, criminal penalty, license suspension and fi nally licensing revoca-

tion. Sparrow makes a similar point that administrative agencies often

jump to the conclusion that they need the biggest weapon, when in fact

a more eff ective solution would be increased vigilance in surveying the

problem.

Warning letters sent by the regulator may serve to reduce the amount

of damage as soon as a complaint is made, and fi t well into the regulatory

pyramid of proportionate responses. There is precedent for this type of

approach in the United Kingdom Communications Act 2003. Under this

regime, the Offi ce of Communications (Ofcom) may send a warning letter

which specifi es a period during which the person notifi ed may make rep-

resentations, bring itself into compliance or remedy the consequences of


the contravention. Warning letters may also avoid the vagueness problem

that has been identifi ed with a principle-based system, as they alert the

regulated that they are potentially in violation of the principles (Halpern

and Puri, 2008, p. 212).

Following the warning letter, if appropriate compliance steps have not

been taken, the next step up in the pyramid is an enforcement notifi ca-

tion. This notifi cation may impose a requirement to take specifi c steps to

comply, and may be enforced by injunctive relief. Finally, penalties may

be imposed for contravention of the enforcement notifi cation. A pyramid

approach is built into the penalty section by explicit reference to the prior

failure to comply with lower remedies such as the warning letters. The

restorative justice approach should not be construed as soft on enforce-

ment. To the contrary, the restorative justice enforcement pyramid works

because, at the highest point, there is enforcement of the most serious

criminal penalties which apply after lower- level approaches have failed.

Part of the reason for the eff ectiveness and success of cooperative meas-

ures at the base of the pyramid is the threat of escalation.

6.3 CONCLUSION

The intention of this chapter has been to provide a framework that can

form the basis for new approaches to the regulation of telecommunication.

Three points are fundamental in this regard and have implications for the

regulation of telecommunications.

First, governments and regulators ought to focus on the prevention of

harms and the serving of human needs, in the order of their priority, when

they make decisions about the allocation of scare public resources.

Second, it is important to emphasize that regulation of competitive

markets will encompass risk management and social regulation. Robust

regulation that ensures proper risk management and transparency is

essential in a competitive economy. In addition, the new rules will target

anticompetitive behavior in line with other competitive industries that fall

within the Competition Act.

Third, the regulatory state should off er multiple models that refl ect the

diff erent needs and interests at stake within both the ex ante and the ex post

paradigms. There is growing recognition in legal circles that any sophisti-

cated system ought to off er a full range of models based on the seriousness

and economics of the case. In the fi eld of telecommunications regulation,

this insight requires: (1) that a menu of regulatory approaches be acknowl-

edged and available to regulators; (2) that principles be articulated for

making choices within these approaches; and (3) that these approaches


be applied with an alert sensitivity to subject matter, the human needs at

stake and the economics of a given case.

Each sector ultimately is enforced ex post if there is a breach of the

rules or principles. The choice between systems relates to types of rules

and principles on a continuum. Prior approval is suitable in cases where

serious harm may be prevented, whereas principles are generally more

suitable for economic regulation in the market circumstances now prevail-

ing in Canadian telecommunications.

NOTES

1. An earlier version of this chapter was published in the Canadian Business Law Journal. 2. Robertson, G. (2008). 3. Since the passage of the Telecommunications Act in 1993, the CRTC has issued a series

of forbearance rulings with the result that most telecommunications services can now be off ered without prior approval. Economic regulation, and the accompanying require-ment for prior approval of rates, is now largely confi ned to local telephone service in remote and rural areas and wholesale services off ered to competitors. However, it should be emphasized that even where the CRTC has forborne, such forbearance is typically partial and conditional, in one way or another, with the result that the CRTC still maintains some form of regulatory oversight over virtually all telecommunications services in Canada, albeit that the CRTC does not impose price controls on forborne services. Consequently, even after substantial forbearance from economic regulation (or because of it), it is necessary and important to consider new and better ways to approach regulation in the telecommunications sector.

4. According to Posner (2003): ‘The movement, which may be said to have begun in the 1970’s in the airline industry, though there were earlier anticipations, coincided with increased regulation of health, safety, and the labor markets, so that to speak of “dereg-ulation” in the large is misleading; regulation has changed, rather than diminished.’

5. Public risk perceptions may be a poor guide for policy- making (Cross, 1998) and public participation may lead to overemphasis on some issues and underemphasis on others (Green, 1997).

6. There is a school of political science that attempts to build a political theory on the basis of human need rankings. Bay (1981) argues that the only legitimate justifi cation of gov-ernment is its ability to serve human needs in the order of their priority as established though psychological research. See also Manzer (1974).

7. ‘The eff ort to constrain, I argue, is more likely to produce distortions than to bring about a reasonable simulacrum of competitive pricing and output. This is primarily because of information and incentive problems of regulators and because of eff orts by the regulated fi rms to neutralize regulation or to bend it to their advantage’ (Posner, 1999, p. vii).

8. We restrict the defi nition of minority to that used in human rights legislation, which recognizes historically disadvantaged minority groups who may lack power within a democratic majority.

9. Such ex ante rules might focus, inter alia, on key size, keyboard layout, and LCD screen size and design, but would fall out of date very rapidly as technology evolved and accordingly would have every limited benefi ts.

10. Braithwaite, J. (1985).11. Residential telephone service is one such example. The service remains rate- regulated in

much of North America notwithstanding the presence of interplatform competition among


cable, wireless, broadband and incumbent telephone carriers. Where local telephone service is not forborne from regulation, it is nonetheless often subject to both inter- and intramodal competition and this fact should inform choices about regulatory models.

12. The FCC has exhibited a basic humility about the ability of prescriptive rules to antici-pate, codify and respond to network neutrality issues. Instead of developing a frame-work of prescriptive rules, the FCC has adopted a set of basic principles to address network neutrality issues.

13. Occupational Health and Safety Act, O. Reg. 213/91, s. 26.3(4), amended to O. Reg. 527/00.

REFERENCES

Archibald T., K. Jull and K. Roach (2004 – updated annually). Regulatory and Corporate Liability: From Due Diligence to Risk Management. Ontario: Canada Law Book, a Division of the Cartwright Group, Chapter 1, pp. 1–21.

Ayres, I., and J. Braithwaite (1992). Responsive Regulation: Transcending the Deregulation Debate. New York: Oxford University Press.

Baldwin, R. (1990). Why rules don’t work. Modern Law Review 53, 321.Bar- Ilan, Y.A. (2007). Justice: when do we decide? Connecticut Law Review 39,

923–32.Bay, C. (1981). Strategies of Political Emancipation. Notre Dame, IN: University

of Notre Dame Press.Braithwaite, J. (1985). To Punish or Persuade: Enforcement of Coal Mining Safety.

Albany, NY: State University of New York Press.Braithwaite, J. (2002). Rules and Principles: A Theory of Legal Certainty.

Australian Journal of Legal Philosophy, 27, 47–82.Breyer, S. (1993). Breaking the Vicious Circle: Toward Eff ective Risk Regulation.

Cambridge, MA: Harvard University Press.Cross, F.B. (1998). Facts and values in risk assessment. Reliability Engineering &

System Safety.Cunningham, L. (2006). Principles and rules in public and professional securities law

enforcement: a comparative US–Canada inquiry. Research study commissioned by the Task Force to Modernize Securities Legislation in Canada.

Dembo, R.S. and A. Freeman (1998). Seeing Tomorrow: Weighing Financial Risk in Everyday Life. Toronto: McClelland & Stewart.

Ehrlich, I. and R. Posner (1974). ‘An economic analysis of legal rulemaking’. Journal of Legal Studies, 3 (1), 257–72.

Ford, C. (2008). New governance, compliance, and principles- based securities regulation. American Business Law Journal, 45 (1), 1–60.

Globe and Mail (2007). ‘Sarbanes, Oxley defend law’s strict measures’, 21 July, p. B5.

Goebel, J. (1993). Rules and standards: a critique of two critical theorists. Duquesne Law Review, 31, 51–68.

Government of Canada (2006). Canadian Radio- television and Telecommunications Commission, Discussion paper, 15 August, submitted to the Telecommunications Policy Review.

Governor in Council (2006). Order issuing a direction to the CRTC on implement-ing the Canadian telecommunications policy objectives. P.C. 2006- 1534, 14 December.


Green, A.J. (1997). Public participation and environmental policy outcomes. Canadian Public Policy Analysis, 22 (4), 234–435.

Halpern, P. and P. Puri (2008). Refl ections on the recommendations of the Task Force to modernize Securities Legislation in Canada: a retail investor perspec-tive. Canadian Business Law Journal, 46, 199–212.

Hill, S. (1969). A foundation for developing risk management learning strategies in the public service. CCMD Roundtable on Risk Management (Canada). Retrieved from http://www.ccmd- ccg.gc.ca/research/publications/pdfs/risk_mgnt_rt_e.pdf.

Janisch, H. (2005). Response of Hudson Janisch to the CRTC’s Discussion Paper, Annex C, Reply Comments of TELUS Communications, Telecommunications Policy Review, 15 August.

Manzer, R. (1974). Canada: A Socio- Political Report. Toronto: McGraw- Hill Ryerson.

Ministry of Transportation (2008), Ontario to limit truck speed. Ontario Government, 16 June. Retrieved from http://ogov.newswire.ca/ontario/GPOE/2008/06/16/c4593.html?lmatch=&lang=_e.html.

Mullan, D. (2005). Tribunals imitating courts: foolish fl attery or sound policy? Dalhousie Law Journal, 28, 1.

Posner, R. (1999). Natural Monopoly and its Regulation. Washington, DC: Cato Institute, p. vii.

Posner, R. (2003). The Economic Analysis of Law, 6th edn. New York: Aspen Publishers.

Powell, D. (2000). Risk- based regulatory responses in global food trade: a case study of Guatemalan raspberry imports into the United States and Canada, 1996–1998. In G. Bruce Doern and T. Reed (eds), Risky Business: Canada’s Changing Science- Based Policy and Regulatory Regime (p. 134). Toronto: University of Toronto Press.

Robertson, G. (2008). Canada’s 911 emergency. Globe and Mail, 20 December, p. A1.

Robertson, G. (2009). Cellphones ordered to fi x 911 systems to save lives. Globe and Mail, 7 January, p. A1.

Sharpe, R.J. (1998 – updated annually). Injunctions and Specifi c Performance, 2nd edn. Ontario: Canada Law Book, a Division of the Cartwright Group Ltd.

Slovic, P. (1999). Trust, emotion, sex, politics and science: surveying the risk- assessment battlefi eld. Risk Analysis, 19(4), 689.

Sparrow, M. (2008). The Character of Harms: Operational Challenges in Control. New York: Cambridge University Press.

Sunstein, S. (1995). Problems with rules. California Law Review, 83(4), 953.Tremblay, A., G. Sinclair and H. Intven (2006). The Telecommunications Policy

Review Panel Final Report. Retrieved from http://www.telecomreview.ca/eic/site/tprp- gecrt.nsf/vwapj/report_e.pdf/$fi le/report_e.pdf.

149

7. The transformation of telecoms industry structure: an event study

Olaf Rieck

7.1 INTRODUCTION

The telecommunications industry is currently going through yet another

phase of transformation. This transformation does not merely change the

way a set of given services is off ered. Rather, we observe that technologi-

cal changes and service innovations are fundamentally altering peoples’

communication patterns. Examples include the shift of the telephone

network from being a physical network to being a (potentially converged)

Internet application, the shift from a focus on connectivity to a focus on

content, and the shift from dyadic communication relationships to social

networking.

In this emerging new telecommunications paradigm, traditional busi-

ness models are increasingly replaced by alternative models. Players

that traditionally controlled the telecoms value chain are rapidly losing

ground against new entrants, some of which entered from ‘outside’ the

traditional telecommunications service industry or did not even exist

at the time when the 1996 Telecommunications Act was drafted. For

example, Google emerged as world market leader for online searches and

has thereby established itself as a signifi cant player in the telecoms value

chain. Google is also – directly or through investments – involved in the

rollout of Internet access infrastructure and the provision of content. Not

surprisingly, in September 2007 the combined market capitalization of

Google and its close competitor, Yahoo!, surpassed the market capitaliza-

tion of AT&T.

While there exist a number of studies that analyze telecommunica-

tions carriers’ performance in their traditional line of business, only a few

studies (Rieck and Doan, 2007) have attempted to assess quantitatively

the new strategic directions that fi rms take in this changing environment.

This chapter studies activities (conducted by telecom carriers or by other

related fi rms) that aim at vertically integrating diff erent segments of the

telecom value chain. It then looks at the implication of these activities


on telecom carrier performance. I am interested in fi nding out to what

extent telecom carriers are threatened by other fi rms’ initiatives targeted

at ‘invading’ their turf, and how eff ective they are in countering such

threats.

This chapter is organized as follows. Section 7.2 will provide a brief

introduction into the research background and method, and the data col-

lection. Section 7.3 will discuss the empirical results. Section 7.4 will sum-

marize and conclude.

7.2 RESEARCH BACKGROUND AND METHOD

7.2.1 The Telecommunications Value Chain

Since this chapter is fi rst and foremost concerned with changes in the

telecommunications value chain, it is important to begin by providing a

simplifi ed model of this value chain. Value chain models can generally not

be assessed in terms of whether they are right or wrong, but rather in terms

of whether they are useful in structuring a given problem or illustrating a

point. For this chapter I we will use a modifi ed version of the layer model

discussed by Fransman (2002). This model breaks the industry into fi ve

layers, which are depicted in Figure 7.1.

Firms in Layer 1 develop and manufacture telecommunications equip-

ment, including networking and customer premises equipment. Examples

for such companies are CISCO, Nortel, Nokia, Motorola, Sony- Ericsson

and Apple (due to their widely celebrated iPhone). Layer 2 includes some

of the key elements of the traditional telecommunications operator busi-

ness, namely the ownership of network facilities. These include both cables

and licenses to spectrum. Companies that dominate this layer are the likes

Layer 5 Content Provision

Layer 4 Content Integration

Layer 3 Network Operation

Layer 2 Facilities Provision

Layer 1 Equipment Manufacturing

Figure 7.1 Telecommunications value chain

The transformation of telecoms industry structure 151

of Verizon, AT&T, Deutsch Telekom and SingTel, but also cable opera-

tors, or the mobile carriers such as Vodafone, Orange and M1. Layer 3

comprises of all activities related to the operation of networks. These may

or may not involve the companies owning the underlying infrastructure.

The separation of facility ownership from service operation is common

in fi xed line telecommunications, but may also become a more common

model in the mobile communications value chain. After all, the regula-

tory foundations were laid as recently as March 2008 – the conditions for

the Federal Communications Commission (FCC)’s 700 MHz frequency

band auctions include an open access rule, which allows facilitates market

entry of purely ‘service- based’ mobile operators. Layer 4 is concerned

with content integration (Fransman refers to this layer as ‘Navigation

and Middleware’) and is dominated by Google, followed by Yahoo! and

Microsoft. Finally, the top layer encompasses all activities involving the

provision of content and applications. This can be, for example, the pro-

duction of movies or news, the provision of online brokerage services,

online travel brokers, or the development of games. Notable companies in

this layer are Time- Warner, BMG, NBC and Travelocity.

It is important to note that, while most companies originated pre-

dominantly from one of these layers, many of them have tried to integrate

vertically into activities in diff erent upstream or downstream layers of

the telecommunications value chain. Again, to give some example, Time

Warner acquired AOL, and Deutsche Telekom bought (and later sold) the

online newspaper Bild.de.

Attempts to extend market power in the telecom value chain are not

necessarily limited to the acquisition of ownership stakes. Rather, they

often take the form of organic entry. Some of the most discussed examples

of the past years are Apple’s launch of the iPhone, which threatens various

traditional players in the value chain. Besides eff ectively becoming a new

player in the hand phone equipment market (Layer 1), Apple has managed

to (technically) tie its handset with its online music and video store iTunes

(Layer 4), and has even attempted to negotiate cuts in the mobile opera-

tors’ revenues (Layer 2) in exchange for letting them exclusively sell their

iPhones. Apple has also secured a number of content deals with content

providers, thereby extending its market power in the content integration

layer (Layer 5).

Nokia and Sony- Ericsson have followed Apple’s lead by launching their

own content platforms. At the same time, Google has been working on

invading the handset market, fi rst by working on a new operating system

for Smartphones (Android), and later by launching its own Smartphone

(‘G1’). Google has also exerted its infl uence extending beyond content

integration by fi rst lobbying for an open access rule in the US 700 MHz


spectrum auctions (such as to open the spectrum to resale if licenses were

auctioned off at a price of at least US$4.6 billion) and then bidding the

price up this threshold such as to trigger the open access rule.

Traditional telecommunications carriers, while under threat from all

sides to be reduced to ‘bit- pipes’, have tried to counter the threats by also

engaging in activities in vertically related markets. These include initiatives

like joining the open handset alliance, the development of mobile portals,

or striking deals with content providers such as to strengthen their posi-

tion in the content integration layer.

In light of all of these activities, the questions arising are precisely how

the structure of telecommunications markets is likely to evolve, which

companies will be able to extend their reach and power, and whether there

are companies whose role may be reduced in future. In my view, the best

places to look for an answer to these questions are the (forward- looking)

fi nancial markets, which do react to the changes described above and

refl ect their evaluation in the market stock prices of the involved compa-

nies. I hence conducted a set of event studies to assess the impact of various

strategic initiatives (‘events’) on the valuation of telecom operators.

In order to get a comprehensive view of changes in the telecom value

chain, I studied a broad range of events where telecom industry players

entered vertically related segments of the telecom value chain. I classifi ed

these events into ‘vertical integration through M&A’ (mergers and acqui-

sitions) and ‘vertical integration through organic entry’, and conducted

two separate studies for these two classes of events. I felt that the charac-

teristics of M&A and organic entry events are too diff erent to integrate

them in one joint study.

M&A have been frequent occurrences in the telecommunications indus-

try ever since the breakup of the telecom monopolies. I was therefore able

to collect data on a relatively large number of M&A events that constitute

some degree of vertical integration and of M&A events that occurred

within one layer of the value chain (the latter is typically related to geo-

graphical expansion). Given the amount of information I had, I found

it useful to compare and benchmark the performance of vertical M&A

against that of intra- layer M&A.

Besides, a signifi cant incidence of organic entry is still relatively rare.

Organic entry tends to be somewhat less well defi ned than M&A and

harder to pinpoint to a particular point in time. I therefore needed to

restrict my study to a few pivotal events in this category, which makes

the treatment of organic entry rather exploratory in nature. Even rarer is

organic entry of telecom carriers into markets that are in the same layer of

the telecom value chain. As a result, I was not able to compare and bench-

mark vertical organic entry against intra- layer organic entry. I therefore


proceeded by treating M&A and organic entry in two separate event

studies, as detailed below.

7.2.2 The Event Study Method

There have been numerous event studies undertaken in a variety of areas,

foremost in the fi elds of accounting, fi nance and strategic management. For

instance, Subramani and Walden (2001) study the returns to shareholders

in fi rms engaging in e- commerce. Johnson et al. (2005) examine the impact

of ratings of board directors by the business press on stockholders’ wealth.

Another area in which event studies have been widely used is in the evalu-

ation of M&A. For instance, Wilcox et al. (2001) analyze M&A events in

the telecommunications industry by testing the impact of near diversifi ca-

tion, far diversifi cation, and the size of the fi rms on the shareholder value.

Uhlenbruck et al. (2006) focus on acquisitions of Internet fi rms and the

potential for the transfer of scarce resources in a resource- based view.

Finally, Rieck and Doan (2007) study the value eff ects of diff erent types

of M&As in the telecommunications industry. However, event studies also

lend themselves well to evaluating the impact of other kind of corporate

activities, as long as they can be traced to some kind of singular event such

as a corporate announcement or a service launch.

The event study method is based on the assumption that capital markets

are effi cient. Based on this assumption, it allows estimation of the impact

of new information on anticipated future profi ts of the fi rms. If informa-

tion communicated to the market contains any useful and surprising

content, an abnormal return – which is defi ned as the diff erence between

the actual post- event return and the return expected in the absence of the

event – will occur (MacKinlay, 1997).

The assumption of effi cient capital markets has also given rise to

criticism of the event study method, particularly in times of stock market

bubbles and busts. In response to these valid concerns I would like

to note two points. First, the alternative to event studies in the realm

of quantitative performance analysis is the use of accountancy data.

Confi dence in accountancy data has been periodically shaken (recall the

Enron scandal), as has the confi dence in the rationality of stock markets.

McWilliams and Siegel (1997) compare the event study methods which

use accounting measures and state that: ‘The event study method has

be come popular because it obviates the need to analyze accounting- based

measures of profi t, which have been criticized because they are often not

very good indicators of the true performance of fi rms’. Second, the event

study method looks at stock returns net of the general market trend in

an industry. Hence, the presence of a general euphoria or panic will not


aff ect the results of this method. In the event study method, what matters

is the stock performance relative to the industry performance. Overall it is

expected that event studies will con tinue to be a valuable and widely used

tool in economics and fi nance.

According to MacKinlay (1997) and Rieck and Doan (2007) an event

study can be roughly categorized into the following fi ve steps:

identifying of the events of interest and defi ning the event window ●

size;

selection of the sample set of fi rms to include in the analysis; ●

prediction of a ‘normal’ return during the event window in the ●

absence of the event;

calculation of the abnormal return within the event window, where ●

the abnormal return is defi ned as the diff erence between the actual

and predicted returns;

testing whether the abnormal return is statistically diff erent from ●

zero.

The events of interest in this study are of two kinds:

1. M&A announcements of major telecommunication companies listed

in the US or European stock exchanges. These may or may not

involve some degree of vertical integration in the telecom value chain.

Looking at both types of M&A (that is, with and without verti-

cal integration) allows us to contrast empirically these two types of

strategies.

2. Any events that constitute an instance of organic vertical integration

in the telecommunications value chain. These include any new ven-

tures that do not fall into the category of M&A activities. Note that

here we are looking not only at the telecommunications carriers’ own

strategic activities to integrate within the value chain. Rather, in the

context of this study an event is defi ned as any activity that aims at

extending a fi rm’s position in the telecom value chain. However, also

note that for all of these events I study only their value eff ects on the

telecommunications carriers.

Defi ning an event window means fi xing the time period before and after

a focal event, within which any abnormal return is attributed to this focal

event. This study uses three symmetric event windows: three- day (one day

prior to the event day and one day after the event day), fi ve- day (- 2; +2)

and 11- day (- 10; +10) event windows. These window lengths are appropri-

ate to capture any news that might have leaked shortly before the offi cial


announcement was made and also to consider any short- term stock price

reactions linked to the event after the announcement.

The estimation window is the control period which precedes the event

period. In this study, the estimation window (denoted as L1) for all

events ends 12 days before the event and extends back to 120 days prior to

the event date. Estimation periods generally end before the event of inter-

est so that the returns during the event window will not infl uence the model

parameters. Figure 7.2 illustrates the situation for the case of a 5- day event

window.

The companies examined in this study are telecommunications carri-

ers that are listed on (at least) one of the major US or European stock

exchanges. For this study of M&A events, the population consists of all

M&A announcements released by these telecommunications companies.

For the study of other (non- M&A) events, the population consists of a

set of initiatives that are either announced by the telecom carriers them-

selves, or by other companies, but directly aff ect the telecommunications

companies one way or the other. In practice, the population list of M&A

announcements is generated by searching for specifi c terms in the titles of

media and news releases.

If the impact of M&A announcements on stock returns is to be exam-

ined, a measure of what shall be the ‘normal’ return for the given stock

is required; 120 days of histori cal stock data will be used for each event.

These 120 days are enough to calculate valid estimators needed for the

event study model (MacKinlay, 1997).

The event study methodology relates the historical stock data in the

estimation window to the market index. In the case of this study, the Dow

Jones Titans Telecommunications Sector Index was used as the market

index, as detailed below. To predict each fi rm’s market model, daily returns

were used to estimate a regression equation over the estimation period. The

underlying securities are assumed to be independently and jointly normally

distributed and shall be identically distributed through time (MacKinlay,

1997). For any company i, the market model is specifi ed as:

Rit 5 ai 1 biRmt 1 eit (7.1)

–2 +2–12–120

L2: event windowL1: estimation window

days

Figure 7.2 Estimation and event window on a timeline


where Rit is the return of security i and Rmt is the rate of return of the

market portfolio in period t. eit is the zero- mean disturbance term. ai and

bi are fi rm specifi c parameters of the market model. The market model

assumes that in the absence of the event, the relationship between the

returns of fi rm i and returns of the market index remains unchanged and

the expected value of the disturbance term eit is zero. Using this approach

the resulting regression coeffi cients and the fi rm’s actual daily returns were

then used to compute abnormal returns for each fi rm over each day of the

event window period. The sample abnormal return ARit on the event day

t is calculated for the ith fi rm by subtracting the predicted return of the

market model from its observed return:

ARit 5 Rit 2 (ai 1 biRmt) (7.2)

where the coeffi cients ai and bi are ordinary least squares estimates of ai

and bi. The Cumulative Abnormal Return (CAR) for fi rm i over the event

period t1 to t2 is then calculated as follows:

CARt(t1, t2) 5 a

t2

t5t1

ARit (7.3)

where (t1, t2) is the event window interval; and all other terms as previ-

ously defi ned. The abnormal returns represent the extent to which actual

realized returns on any of the event days deviate from the expected returns

based on the estimated fi rm- specifi c market model. In this sense, the abnor-

mal returns can be seen as prediction errors (eit). MacKinlay (1997) then

goes on developing the methods to test whether or not the computed CAR

is signifi cantly diff erent from zero. The respective formulas are omitted at

this point, but can be found in many of the papers quoted above.

7.2.3 Data

The two data inputs required for the M&A event- study model are the

events themselves (in this case M&A announcements and other non- M&A

events), and historical stock price data (security prices and the reference

index). They were both gathered from ABI/Inform and Yahoo.com.

In order to explore the eff ects of the selected events on stock prices,

this research limits its scope to companies that are either listed on one of

the major European stock exchanges (London, Paris, Frankfurt, Madrid,

Amsterdam) or on a US stock exchange (NYSE, NASDAQ). The reason

is that most of the telecommunication operators are listed on at least one

of these stock exchanges. It is generally known that these exchanges inspire

high public confi dence due to the high standards and listing requirements.


Thus, using only major exchanges also increases the price stability of the

securities – an important prerequisite for the event- study method.

Diversifi cation through mergers and acquisitions

For my M&A analysis I collected a sample of M&A completed by publicly

traded acquirers between 1998 and 2006. Telecommunication operators

were selected from the Thomson One Banker database indexes by search-

ing for companies within the following industries: companies with primary

SIC Code 4813 (Telephone Communications except Radiotelephone),

4812 (Radiotelephone Communications) or 4842 (Cable and Other Pay

Television Services). The results were limited to companies that were (at

the time of the event) listed on a major US or European stock exchange.

Next, with a list of 56 potential acquirers, the M&A events which

are associated with every one of these carriers were separately retrieved

from the Highbeam™ Research database.1 The following search on

Highbeam™ using a set of relevant search terms was performed and the

earliest dates announcing the event were recorded:

Search in article title only: ●

{company name from list} AND buy OR acquires OR bid OR merge

OR takeover OR merger OR acquisition OR buys OR merges OR

merging OR acquiring

Search in the following sources: ●

{news sources from the following list: Business Wire, Associated Press,

PR Newswire, Reuters, Wall Street Journal, Washington Post}

Dates: ●

1 January 1998 and 31 March 2008

Events that were identifi ed using these criteria were consolidated into

a master list with duplicates removed. The preliminary sample frame had

512 M&A events. Out of this, an acquisition that resulted in a controlling

stake for the acquirer, that is, a stake greater than 50 percent, was chosen.

This means that acquisitions giving the acquirer lower than 50 percent in

stakes are by defi nition not considered as acquisitions. It is true that that

in practice acquirers may de facto control the target even though they hold

less than 50 percent of the shares. Moreover, it is also true that owning

only a minor stake, say 10 percent of the shares, could hardly result in a

signifi cant degree of control of the target. Determining whether or not an

acquisition resulted in a controlling stake is diffi cult without using a great

deal of insider knowledge and subjectivity. I therefore chose to apply the

50 percent rule outlined above, acknowledging the practical limitations of

our approach.


Moreover, only M&A announcements containing accurate and detailed

information about date of announcement, partner and transaction value

were included. Further, only those events with at least 120 days of historic

stock data2 available were selected. Last, to avoid possible confound-

ing eff ects within the event window, a number of M&A events were

selectively omitted. Excluded events are those that coincided with other

major fi rm- specifi c events that might aff ect the stock price, such as earn-

ings alliance announcements, earnings, large investment decisions or new

product introduction (McWilliams and Siegel, 1997). Confounding events

were identifi ed using the Highbeam™ Research function. It displays all

company news within a four- day range of the specifi c date. After meeting

all these criteria, the fi nal sample contained 88 M&A announcements of

37 telecommunication fi rms. Once the M&A announcements were isolated

and the event window was defi ned, the ‘normal’ (expected) returns for that

window needed to be estimated. This was done by using historical stock

price data (adjusted closing prices) for all acquiring companies listed in

the master list. For each event 132 days of historic stock data before the

event date and ten days of stock data after the event date were downloaded

through either the Center of Research on Security Prices database (CRSP)

or the Yahoo! Finance database.

To measure the direction of diversifi cation of the M&As, this study

employs the SIC3 classifi cation as conducted by Berry (1971) and Ferris

and Park (2001). Many industrial organization studies have used objec-

tive measures based on standard industrial classifi cation (SIC) counts

to capture the aspect of diversifi cation (Ramanujam and Varadarajan,

1989). The fi rst number assigns a product to a very broad category. Each

subsequent number distinguishes the product at a progressively fi ner

level. The SIC classifi cation has been widely used among economists to

determine in which industry segments the company is operating. As all

acquirers in this study are listed telecommunications companies, they all

operate with the two- digit 48xx SIC code. The acquirer’s strategy can be

determined by comparing the SIC codes of the acquirer and the target.

Telecommunications M&A occurring solely within the 48xx SIC code

(that is, both acquiring and target fi rms) are termed non- conglomerate

mergers. M&A where the target has a SIC code other than 48xx are clas-

sifi ed as conglomerate mergers (Ferris and Park, 2001; Ramanujam and

Varadarajan, 1989). Conglomerate mergers include all mergers across dif-

ferent parts of the value chain.

Organic diversifi cation events

The data collection for relevant organic diversifi cation events was con-

ducted in a similar fashion as for the M&A events; however, it was


somewhat less systematic and more exploratory. The reason is that in the

case of M&A it is intuitively clear what constitutes a ‘signifi cant’ event

(that is, the acquirer purchases a signifi cant number of that results in a

voting majority). If a company organically enters a part of the telecom

value chain, it is not immediately clear what criteria to apply in order to

select only signifi cant events, that is, events that are signifi cant enough to

make a noticeable impact on the industry structure. In this chapter, I will

report the impact of a few such events, which I consider intuitively as sig-

nifi cant. These events are:

1. Apple’s launch of the iPhone (and the signing of exclusive contract

deals).

2. Nokia’s announcement of OVI.

3. Sony- Ericsson’s announcement of a major content deal with three of

the four major music labels.

4. Google’s announcement of Android.

5. Surpassing of the US$4.6 billion mark in the 700MHz license auc-

tions, which triggered the open access rules.

6. Cingular Wireless, Verizon and Sprint announcing their music down-

load portals.

Since I expected the respective announcements to be somewhat less

of an immediate surprise, I chose a 21- day event window. I studied the

impact of the six events on the aff ected telecommunications carriers’ eval-

uation. In other words, I studied the impact of the iPhone announcement

in the UK on the Telefonica (O2) valuation, as well as on the evaluation

on its competitors. I did not however consider the impact on the valuation

of Apple Inc. Similarly, I studied the impact of the OVI announcement

on the aff ected telecom operators (note that most large operators are

trying to off er similar services, so OVI is a clear threat to their business

expansion into the content integration market), but I did not study the

impact of the OVI announcement on the Nokia share price. Beyond that,

I followed the same data collection procedure as outlined in the case of

the M&As.

7.3 EMPIRICAL RESULTS

Table 7.1 presents a summary of the results from the event study of M&A

and shows the statistical values of fi rm diversifi cation strategy, that is,

conglomerate and non- conglomerate mergers.

The fi rst two columns of Table 7.1 present coeffi cients and signifi cances


for the ‘additive event- window abnormal returns’4 in the respective event

window (also referred to as ‘mean cumulative abnormal return’ – CAR5).

The abnormal returns are calculated using the market model estimated

from 132 to 12 trading days prior to the event announcements. The Mean

CARs (given in parentheses) represent the cumulative market model-

adjusted abnormal returns over the relevant event window. The Z- statistics

for the (- 1,+1) and (- 2,+2) event windows are based on the standardized

abnormal return method according to Patell (1976).

The fi rst row of Table 7.1 reports the results for the complete sample.

There is signifi cant support for the hypothesis that M&A activities will,

in general, have a positive impact on telecommunication fi rms partici-

pating in these activities. The mean CAR for both windows is approxi-

mately +0.85 percent and signifi cant at the 5 percent level. If we break

the set of M&A events up into conglomerate mergers (that is, mergers

within a segment of the value chain), non- conglomerate mergers (that

is, diversifi cation across segments of the value chain), we get the follow-

ing results: non- conglomerate mergers will generate positive abnormal

returns. CARs of +0.63 percent and +0.61 percent are reported for

both windows, respectively, both which are weakly signifi cant at the 10

percent level.

The CAR for the conglomerate mergers is twice as high (+1.29

percent) and is signifi cantly positive at the 5 percent level for the (- 1,+1)

window. However, no statistically signifi cant evidence was found in the

(- 2,+2) window. The t- score for the paired t- test that measures the dif-

ference for both CAR means (Δ = NonConglom.CAR - Conglom.CAR)

is negative, but not signifi cant. Hence, there is no signifi cant evidence

Table 7.1 Event study results: diversifi cation through M&As

Event types N Z- value for mean CAR

(- 1,+1) (- 2,+2)

Complete sample 88 2.234*

(0.85%)

1.871*

(0.86%)

Firm diversifi cation

Non- conglomerate merger 58 1.478#

(0.63%)

1.427#

(0.61%)

Conglomerate merger 30 1.764*

(1.29%)

1.219

(1.19%)

Note: The symbols #,*,** and *** denote statistical signifi cance at the 10%, 5%, 1% and 0.1% levels, respectively, using a 1- tail test.


that conglomerate mergers show diff erent abnormal returns than non-

conglomerate mergers.

Table 7.2 shows the CARs related to a number of events that alter the

market structure in the value chain, but are organic in nature (that is, not

related to M&As). While these results are fairly exploratory in nature,

there are a number of interesting trends to be discussed. First, overall tel-

ecommunications carriers appear to be highly threatened by the majority

of changes that currently occur in the value chain. This is indicated by a

highly signifi cant negative CAR of the complete sample of events. When

looking at some individual components, it appears that it is particularly

the ‘outsiders’, such as Google, Nokia and Ericsson, with their attempts to

reach out into diff erent parts of the value chain that impose a threat to the

telecom carriers by lowering their forward- looking valuation.

So far the carriers themselves could do little to convince markets that

they are capable of launching promising activities. Neither the launch of

carriers’ own music download portals, nor the exclusive iPhone deals were

greeted with a great deal of enthusiasm. In both cases, the changes in CAR

were insignifi cant.

Table 7.2 Eff ects of organic diversifi cation on carrier valuation

Sample type N Z- value

for mean CAR

(- 10,+10)

Complete sample 35 - 0.97***

Subclasses of events – eff ects on telco carriers

Outsiders entering content integration

markets

19 - 0.47*

Telcos’ own entry into content

integration market

3 0.05

Google launching

Android

7 - 0.26*

Triggering of open access rule in

spectrum auctions

3 - 0.43***

Telcos’ exclusive iPhone deals 5 0.14

Note: The symbols #, *, ** and *** denote statistical signifi cance at the 10%, 5%, 1% and 0.1% levels, respectively, using a 2- tail test.


7.4 DISCUSSION AND CONCLUSION

The results of the M&A event study are consistent with previous event

studies in showing that M&A in the telecommunication industry generally

result in signifi cant gains in the market values of the acquirer. Therefore,

it can be concluded that the market is generally optimistic with regards to

the potential of telecom carriers to add value in this industry. The highly

competitive marketplace in the telecom sector means that high returns are

no longer guaranteed for big telecom fi rms. Telecommunications networks

have typically high fi xed costs but comparably low marginal costs. As a

result, the potential for economies of scale and scope remains enormous

in this industry. All rival operators are racing to grow fast to reap those

benefi ts. Inves tors may have realized that long- term growth depends on

capital being diverted to productive purposes. However, a reason for cases

where fi rms do not show positive gains after an M&A announcement can

be that it is not always easy for a company to achieve synergies and to reap

scale and scope. High integration costs and diff erences in corporate culture

are reasons why M&A fail to add value. This shows investors’ skepticism

about the likelihood that the acquirer will be able to realize these synergies

required to justify the premium paid (Selden and Colvin, 2003).

Besides engaging in synergistic M&As, telecom operators have also been

engaging in conglomerate mergers, diversifying into diff erent parts of the

value chain. This type of M&As were also found to be wealth- creating.

Furthermore, wherever I looked at incidences of organic carrier diversifi -

cation, the announcements of these activities did not show any signifi cant

impact on fi rm valuation. It appears that forward- looking markets did

not put a great deal of confi dence into the own innovative strengths of

the carriers in my sample when it came to business ventures outside their

traditional core business. Unfortunately for the carriers, in times of reduc-

ing margins in the core business and commoditization of the traditional

network operation business it is precisely those innovative activities in

vertically related markets that could help them regain their competitive

strength. The Japanese and Korean operators provide a good illustration

of this point.

At the same time the telecom operators I studied appeared to be under

threat by ‘outsiders’, whenever these outsiders engaged in those kind

of activities that shift the market power in the value chain. Financial

markets, therefore, appear to support the nation that telecom carriers may

gradually become reduced to ‘bit- pipes’. The carriers may in future need to

accept that they no longer act from a position of control or strength, but

rather from a defensive position. From a regulatory perspective, telecom-

munication fi rms’ market power is increasingly threatened by ‘outsiders’


who penetrate their traditional markets. Thus, telecom regulation may

need to move further away from the focus on ‘carrier regulation’ and

towards a holistic oversight of competition in the entire telecom value

chain.

This chapter has analyzed the fi nancial markets’ reaction to corporate

activities that aim at changing the market structure within the telecom

value chain. While I do appreciate that fi nancial markets often appear to

behave in a less than rational fashion, I also believe that this reaction is

arguably one of the best gauges of future developments at our disposal. I

would also like to remind the reader that parts of this study are explora-

tory in nature. Particularly, my analysis of organic entry is based on a

rather small sample size. As time passes, more companies will try to extend

their position in the telecom value chain, which will allow researchers

to conduct studies based on larger sample sizes. Nevertheless, with this

study I hope to have contributed to an initial understanding of current

fundamental changes in the telecommunications industry. It can be a basis

for further research in this fi eld and might also be helpful for research

on diff erent industries that share a similar structure and conditions to

telecommunications.

NOTES

1. HighBeam™ Research is an online research engine which sorts free, paid and proprietary online articles and databases. It is a tool for serious business, education and personal research; http://www.highbeam.com.

2. They were retrieved through Yahoo! Finance and the CRSP database.3. A standard industrial classifi cation (SIC) code categorizes US business establish-

ments based upon the type of business activity performed at their location. All fi elds of economic activity are included in this system including both manufacturing and non- manufacturing operations. The system is governed by the Offi ce of Statistical Standards.

4. They represent the sum of the mean cumulative abnormal returns over all event window days.

5. Diagrams for the mean cumulative abnormal returns for the complete sample were gen-erated for the fi ve- , ten- and 20- day event windows.

REFERENCES

Berry, C. (1971). Corporate growth and diversifi cation. Journal of Law and Economics, 14(2), 371–83.

Ferris, S. and K. Park (2001). How diff erent is the long- run performance of mergers in the telecommunications industry? SSRN Working Paper.

Fransman, M. (2002). Mapping the evolving telecoms industry: the uses and short-comings of the layer model. Telecommunications Policy, 26(9), 473–83.


Johnson, J., A. Ellstrand, D. Dalton and C. Dalton (2005). The infl uence of the fi nancial press on stockholder wealth. Strategic Management Journal, 26(5), 461–71.

MacKinlay, A.C. (1997). Event studies in economics and fi nance. Journal of Economic Literature, 35(1), 13–39.

McWilliams, A. and D. Siegel (1997). Event studies in management research: theo-retical and empirical issues. Academy of Management Journal, 40(3), 626–57.

Park, M., D. Yang, C. Nam and Y. Ha (2002). Mergers and acquisitions in the telecommunications industry: myths and reality. ETRI Journal, 24(1), 56–68.

Patell, J. (1976). Corporate forecasts of earnings per share and stock price behav-ior. Journal of Accounting Research, 14, 246–76.

Ramanujam, V. and P. Varadarajan (1989). Report on corporate diversifi cation: a synthesis. Strategic Management Journal, 10, 523–51.

Rieck, O. and C.T. Doan (2007). Shareholder wealth eff ects of mergers & acquisi-tions in the telecoms industry. Proceedings of the ITS 18th European Regional Conference, 4–6 September, Istanbul, International Telecommunications Society.

Selden, L. and G. Colvin (2003). M&A needn’t be a loser’s game. Harvard Business Review, 81(3), 70–73.

Subramani, M. and E. Walden (2001). The impact of e- commerce announcements on the market value of fi rms. Information Systems Research, 12(2), 135–54.

Uhlenbruck, K., M. Hitt and M. Semadeni (2006). Market value eff ects of acquisi-tions involving internet fi rms: a resource- based analysis. Strategic Management Journal, 27, 899–913.

Warf, B. (2003). Mergers and acquisitions in the telecommunications industry. Growth and Change, 34(3), 321–44.

Wilcox, H. Dixon, K. Chang and V. Grover (2001). Valuation of mergers and acquisitions in the telecommunications industry: a study on diversifi cation and fi rm size. Information and Management, 38, 459–71.

PART III

Access regulation and performance in the deployment of NGNs: international experiences

167

8. From the pursuit of effi ciency to the pursuit of competition in New Zealand’s evolving telecommunications market1

Bronwyn Howell

8.1 INTRODUCTION

From an economic perspective, effi ciency is the defi ning performance

benchmark for any industry or sector – not least telecommunications.

Consequently, the primary normative objective of law- and policy- making

is the promotion of economic effi ciency (in both its static and dynamic

forms) via the elimination of market ineffi ciencies (Schmalansee, 1981;

Kahn, 1970, 1975).

A minority of economists, and many consumer advocates, propose the

use of law- and policy- making powers principally as a means of achiev-

ing distributional objectives, independent of their eff ects upon total effi -

ciency (for example, Feldstein, 1972a, 1972b). However, in practice it is

extremely diffi cult to achieve desired distributional outcomes through laws

and policies (Schmalansee, 1981), and attempting to do so may well be

counterproductive (Kahn, 1975; Peltzman, 1976). Furthermore, because

distributional objectives are highly subjective, it is very diffi cult to adjudge

the ‘successes’ of any distribution- motivated intervention. By contrast,

effi ciency is an objective measure that provides a useful benchmark for

the economic assessment of law- and policy- making performance. Even

if redistribution is a primary consideration, the Kaldor–Hicks criterion

requires effi ciency gains suffi cient that the winners from a law or policy

change could compensate the losers and still be better off relative to the

status quo (Connolly and Munro, 1999).

Competition is one important means of increasing both static and

dynamic effi ciency; another is regulatory intervention. Although inter-

vention undoubtedly has many potential pitfalls, its use may be justifi ed

in industries where pursuit of competition (in the form of many market


participants) is unlikely to generate greatest total effi ciency in the long

run; for example, natural monopoly, where high fi xed and sunk infra-

structure costs and preservation of long- term incentives to invest in new,

higher- quality technologies that maximize dynamic effi ciency gains may

require protection of a single operator from competitive entry (Alleman

and Rappoport, 2005). In such industries the challenge for governments

is to determine how to allocate responsibility for industry regulation and

governance between two competing forms of intervention, competition

law and industry- specifi c regulation, such that the pursuit of greatest total

welfare ensues in subsequent industry activity.

Competition law can govern interactions to promote increased effi ciency

in most industries, but has limitations. Courts are reactive, responding

only to those cases and those points of law brought before them, and thus

infl uence only a small range of issues that are not necessarily the most

important from a broader effi ciency perspective. Moreover, even when

cases are brought, they are often adjudicated by generalist antitrust judges

who in many cases may lack the industry- specifi c economic knowledge

upon which effi ciency decisions may turn.

Where the underlying economic conditions are suffi ciently diff erent –

as in network industries such as telecommunications – industry- specifi c

regulation off ers advantages. Industry- specifi c regulators in most cases

have the requisite economic knowledge to give due weight to effi ciency

considerations and to balance the tensions between short- term static and

long- term dynamic effi ciencies. Rather than being reactive, they can be

proactive – a power that, when applied appropriately, can lead to increases

in effi ciency. However, the weakness of industry- specifi c regulation is the

risk of capture, leading to the pursuit of other objectives (such as distri-

butions overly favorable to one group, or competition as an end in itself,

rather than a means) over the pursuit of effi ciency.

This leads to the motivating question for this chapter: ‘Can an industry-

specifi c regulatory regime be designed to maximize the relative benefi ts of

industry- specifi c regulation by binding the regulatory institution to pursue

an economic effi ciency objective, whilst simultaneously avoiding the risks

of regulatory capture?’

Examination of New Zealand’s telecommunications sector over the past

20 years provides some interesting insights on this question. Although the

New Zealand government initially prioritized pursuit of economic effi -

ciency in its ‘light- handed’ regime (particularly the preservation of long-

run incentives to invest in new networks and technologies), and attempted

to safeguard its pursuit in subsequent telecommunications sector legisla-

tion, the objective was nonetheless vulnerable to capture by groups with

strong vested interests. However, as the regulator was bound to prioritize


effi ciency, politicians rather than the regulator were the subject of capture

exercises, as they held the power to alter the terms of the regulatory

objective. Consequently, successive governing parties would alter the leg-

islation enacted by their predecessors in order to pursue diff erent sector

objectives.

Thus, even legislation cannot ensure that effi ciency remains a consistent

goal of sector regulation, as long as regulators remain the agents of politi-

cal principals. This suggests that, imperfect though it may be, competition

law overseen by an independent judiciary off ers the best chance of enshrin-

ing the pursuit of effi ciency in industry governance, even in industries that

are normally the focus of industry- specifi c regulation.

The remaining sections of this chapter describe the development of the

regulatory framework for New Zealand’s telecommunications industry,

noting the shift in emphasis from a light- handed regulatory style with

an effi ciency focus ranking long- run dynamic effi ciency considerations

highest, to a more interventionist industry- specifi c regime with a focus on

achieving certain distributional outcomes and fostering competition as

the pre- eminent objective (where greater weight was given to the pursuit

of static effi ciency gains over total effi ciency in the long run). The chapter

concludes by summarizing the pattern of change in New Zealand, and

drawing potential lessons for regulatory design.

8.2 INCEPTION OF NEW ZEALAND’S LIGHT- HANDED, EFFICIENCY- FOCUSED REGIME

New Zealand led the world in light- handed regulation when, from 1984,

the government embarked upon a comprehensive restructuring of the

country’s economy. The government’s clearly articulated objective was

to increase economic effi ciency and create, as Evans et al. (1996, p. 1863)

cited: ‘wherever possible, a competitive environment in which markets can

operate relatively free from subsequent intervention by Government’. As

part of this process, the government enacted the Commerce Act (1986),

which was designed to promote competition for the long- term benefi t

of consumers, and the Telecommunications Act (1987), which avoided

industry- specifi c regulation in favor of generic competition law under the

provisions of the Commerce Act.

Effi ciency considerations were paramount in the choice of institu-

tional mechanisms for industry governance in New Zealand. Government

advisors explicitly identifi ed that regulatory weaknesses such as high

costs, infl exibility and bureaucratic capture could be avoided under a

competition law regime governing all commercial activity (Ministry of


Commerce and The Treasury, 1995; Blanchard, 1995). However, there

were challenges in the enactment of eff ective competition law, not only

as a result of the limitations attendant with non- specialist judges decid-

ing on competition cases in a range of specialized industries, but also due

to New Zealand’s specifi c economic circumstances: small population,

low population density, geographical isolation, challenging terrain, thin

capital markets and historically highly concentrated industries. Almost all

signifi cant industries in New Zealand have three- fi rm concentration ratios

in excess of 90 percent (Arnold et al., 2003), making oligopolistic rather

than perfect competition the most likely pattern of interaction observed in

all industries, not just those with high fi xed and sunk costs.

On balance, the government favored establishment of a single competi-

tion law- based institution to capture institutional scale and scope econo-

mies as well as to ensure consistency in application across industries. The

presumption was that holding a dominant position was acceptable (indeed

it was unavoidable in almost all industries), but that exertion of that

dominant position in such a manner that it harmed the long- term interests

of consumers (as determined under competition law) was illegal. These

arrangements preserved investment incentives for fi rms in industries with

natural monopoly characteristics by permitting dominance (important,

given high investment requirements in these industries, historically thin

capital markets and a small market in the fi rst place), but relied upon com-

petition law to ensure that the behavior of fi rms with dominant positions

did not adversely aff ect the achievement of static and dynamic effi ciency

objectives. However, Part IV of the Commerce Act explicitly provided for

the government to impose price controls in industries where market power

existed, should this be deemed necessary (for example, in the presence of

repeated, proven abuse of a dominant position).

Although ‘lightly’ regulated under competition law, the telecom-

munications sector was far from unregulated. When the state- owned

monopoly incumbent Telecom Corporation of New Zealand Limited

(Telecom) was privatized in 1990, contractual obligations known as the

‘Kiwi Share’ (subsequently the Telecommunications Service Obligation

or TSO) imposed rural–urban universal service and free local calling obli-

gations upon Telecom, and a price cap on residential line rentals. These

terms could be broken only with the express permission of the Minister

of Communications, and even then only where it could be demonstrated

that Telecom was under fi nancial duress. The paramount principle was

that contractual agreement, enforceable under contract law, rather than

overt regulation, subject to high costs and risks of political capture,

off ered the most cost- eff ective means of advancing the pursuit of increased

effi ciency.


The New Zealand arrangements thus involved the government actively

suspending its ‘primary tendencies’ to bow to vested distributional inter-

ests in order to create a set of governance arrangements that would

insulate the sector from politically motivated actions (which would have

diluted the pursuit of effi ciency). Successful implementation would require

the exercise of signifi cant ministerial restraint in order to allow confi dence

in the court- and contract- based regulatory instruments to build up.

8.3 EXPERIENCES UNDER COMPETITION LAW

During the period of light- handed regulation, only two Commerce Act

actions alleging exertion of a dominant position were brought against

Telecom. In both cases, Telecom was found not to have acted anticompeti-

tively. In one case, entrant Clear Communications alleged that Telecom

had acted anticompetitively by including a component to recover the

costs of the universal service obligation in its interconnection prices to

competitors (the Clear case). After three years and three hearings (in the

High Court, Court of Appeal and Privy Council), Telecom was adjudged

to have acted legitimately in adopting Baumol- Willig Effi cient Component

Pricing (ECP). The other case was brought by the Commerce Commission

in 1999, alleging that Telecom’s charging for residential dial- up Internet

calls was anticompetitive. This case was fi nally adjudicated in 2008, and

again found that Telecom had acted as any competitive fi rm would in the

circumstances (Howell, 2007).

Despite multiple claims of ‘failure’, performance data indicate that

New Zealand’s light- handed regime performed, on a range of static and

dynamic effi ciency grounds, at least as well as other industry- specifi c

regimes. The New Zealand residential telephony price index fell by more

than the Organisation for Economic Co- operation and Development

(OECD) average during the 1990s (refl ecting substantial static effi ciency

gains passed on to consumers), and free residential dial- up Internet access

helped the country become a world leader in Internet connection and use

– indicating that dynamic gains from consumer adoption of new applica-

tions were substantial and in excess of those achieved in other jurisdictions

(Howell, 2007). Furthermore, dynamic effi ciency gains from investment in

new generation technologies did not appear to be impaired: Telecom was

one of the OECD’s earliest digital subscriber line (DSL) adopters (January

1999), using a high- speed (2 Mbps) service that was made widely available

(85 percent of the population had access by 2003) and that was priced at

a very low level, taking speed into account. Furthermore, by 2001, new

networks bypassing Telecom’s infrastructure had been deployed in several


business districts and three substantial residential markets, including the

capital city Wellington (Howell, 2003).

Nationally, however, Telecom still held a very large market share, and

was clearly charging prices in excess of marginal cost in some markets

(even if only to recoup universal service obligation costs). In response,

many market participants – principally competitors to Telecom – and

political opponents of privatization strongly advocated the introduction

of industry- specifi c regulation.

8.3.1 Arguments for a Move Away from Competition Law Regime

The justifi cation most commonly cited for a proposed shift to industry-

specifi c regulation was the extent to which Telecom’s prices exceeded mar-

ginal cost – an issue highlighted in the Clear case. Although the fi nal Privy

Council ruling was in Telecom’s favor, the Court of Appeal had found

that any price including a component of monopoly rent (that is, above

marginal cost) was anticompetitive.

The essence of the Court of Appeal decision in the Clear case – and of

most arguments for the introduction of industry- specifi c regulation – was

a distributional rather than an effi ciency concern. The implication of the

Court’s judgment was that Telecom alone should bear the costs of the

social obligations. If the decision had been enforced, there would have

been some static effi ciency gains available to competitors (and potentially

consumers) in low- cost (urban) areas, but with welfare losses in other

higher- cost (rural) areas and longer- term dynamic effi ciency consequences

from likely ineffi cient entry that would be incentivized.

Under the Court of Appeal decision, competitors’ costs would have been

lower than Telecom’s prices in areas where Telecom recovered the costs of

the social obligations (urban areas), making selective entry into these low-

cost areas profi table for competitors but making Telecom’s business fi nan-

cially unsustainable. Profi ts that Telecom was using to subsidize otherwise

unprofi table connections would instead be transferred to competitors, and

Telecom would be forced to raise prices for all its remaining customers to

remain fi nancially viable. In low- cost areas, such price rises would acceler-

ate consumer substitution to lower- price entrants.

Ultimately, Telecom alone would be left servicing all high- cost areas at

higher prices than previously, lowering consumer welfare in those regions;

and entrants would be servicing the low- cost areas, with higher consumer

welfare only if the prices charged were lower than those originally charged

by Telecom. Furthermore, the resulting two- price equilibrium would be

at odds with the distributional intention of the universal service obliga-

tion that was the source of the cost–price dichotomy in the fi rst place.


Moreover, dynamic effi ciency would be potentially lowered as entrants

could profi tably enter in low- cost areas even if their costs exceeded

Telecom’s, as long as their prices fell below Telecom’s (which included

social cost recovery not imposed on entrants). Although such entry would

lead to decreases in effi ciency overall, it was favored with Telecom’s

competitors and opponents of privatization largely because of the distri-

butional consequences. Under the guise of promoting competition (com-

petitive entry) these stakeholders argued that a disproportionate share of

the costs of social obligations fall on the private owners of the formerly

government- owned network, rather than being spread across the industry,

and thereby across all consumers of telecommunications services, irrespec-

tive of their provider, as part of Telecom’s interconnection charge or as a

tax (Armstrong, 2001).

By contrast, the Privy Council (prevailing) decision did not imply

distributional change, and thus preserved existing network investment

incentives and Telecom’s fi nancial sustainability. The Council ruled that

Telecom had priced exactly as a competitive fi rm could be expected to

price its products, based upon its opportunity cost. However, in making

this ruling, the Council enshrined in precedent a pricing rule that could,

under certain circumstances, result in less than effi cient entry and hence

potentially lower product variety than optimal in downstream markets

(short- term dynamic effi ciency; Economides and White, 1995). The deci-

sion was interpreted by some commentators at the time as an example of

how court- based pursuit of ‘competitive’ outcomes might compromise

pursuit of the effi ciency objectives governing the design of the light- handed

regime (Blanchard, 1994a, 1994b, 1995). That is, pursuing competition

(the means) potentially compromised effi ciency (the end). In practice,

however, the decision likely averted potentially more damaging long- term

dynamic effi ciency consequences, and preserved the distributional objec-

tives for consumers inherent in universal service obligations.

To overcome the perceived shortcomings with court- based decision-

making in the telecommunications industry, some proposed not the

adoption of full industry- specifi c regulation, but rather a light- handed

arbitration alternative. The envisioned alternative would sit under the

Commerce Act and the courts, enabling swifter resolution of disputes and

with a wider mandate to consider effi ciency issues not directly part of court

pleadings (Blanchard, 1995).

8.3.2 Political Review of Competition Law Regime

In 1995, the New Zealand Treasury and the Ministry of Commerce led

an inquiry into the proposed alternatives to competition law (Ministry


of Commerce and The Treasury, 1995). Based on the review fi ndings, the

Minister of Communications demonstrated ministerial restraint when

expressing confi dence in the existing system and stating that no changes

would be made. Thus, the pursuit of total economic effi ciency as the pre-

vailing sector objective and the use of ‘light- handed’ regulation to deliver

it was, for the time being, reinforced at the political level (ibid., para. 209),

despite the recognized (but assessed to be small) risk that this might have

on downstream competition.

However, political change (in the 1999 election, which brought a

Labor Party- led coalition government into power) saw the competition

law regime come under review again in 2000. The Labor Party had cam-

paigned on a promise to reform the Commerce Act to tighten controls on

fi rms with a dominant position, and to launch an inquiry into the conduct

of both the telecommunications and electricity industries. The promised

Commission of Inquiry was established soon after the election, with a

brief to address (amongst other issues): alternative means of establishing

interconnection terms and conditions; pricing principles and other terms

and conditions for interconnection; processes applying to interconnection

negotiations, including dispute resolution and enforcement mechanisms;

local loop unbundling; resale of telecommunications services; information

disclosure; the ‘Kiwi Share obligations’; the numbering regime; number

portability and the development of an information economy; and the

eff ect of regulatory regimes on incentives to invest in new technologies.

The Commission of Inquiry appeared to interpret its brief as expressed

in the government policy statement in effi ciency- related terms. ‘Cost-

effi cient’ was presumed to mean that services are produced ‘at the lowest

cost and delivered to consumers at the lowest sustainable price’ (MED,

2000, p. 11) (that is, perfect productive and allocative effi ciency), ‘timely’

to mean ‘the absence of barriers that would impede the implementation

and uptake of innovative services’ (dynamic effi ciency) and ‘ongoing’ to

mean that ‘regulation should be forward- looking, robust, durable and

consistent over time, and not sacrifi ce long term [sic] gains for short- term

considerations’ (the trade- off between dynamic and static effi ciency). In

addition, the Commission took ‘fair and equitable’ to mean ‘the way in

which services are provided, the conduct of the industry players and their

interactions’, suggesting weight would be given to competitor equity as

well as consumer welfare (and effi ciency).

However, because the government’s policy statement imposed multiple

and confl icting objectives, the Commission of Inquiry was clearly unable

to deliver on all of them. Both static and dynamic effi ciency objectives were

to be addressed simultaneously, with no guidance given as to which should

take priority. Moreover, the policy statement required both effi ciency and


distributional objectives also to be simultaneously satisfi ed. Without clear

prioritization criteria, the policy statement left considerable scope for the

Commission of Inquiry itself to be captured by vested interests.

Unsurprisingly given these confl icting instructions and the strongly

polarized views held by industry participants and political agents, the

Commission of Inquiry’s fi ndings were focused on effi ciency in some

instances, and not in others. For example, its report considered dynamic

effi ciency in rejecting local loop unbundling. Echoing the Court of Appeal

decision, the Commission also considered effi ciency in concluding that

Telecom’s prices still included elements of prices above cost (and thus was

not perfectly allocatively effi cient). However, in proposing a solution to

the perceived failure of light- handed regulation, the Commission did not

consider effi ciency: the Commission proposed that the government estab-

lish an industry- specifi c regulator to set total service long- run incremental

cost (TSLRIC)- based prices for a variety of Telecom’s fi xed- line services

(excluding value- added services such as asymmetric digital subscriber line

– ADSL), but ignored dynamic effi ciency issues in its cost–benefi t analysis

of a new regulatory agency. Furthermore, distribution considerations –

rather than effi ciency considerations – prevailed in the Commission of

Inquiry’s recommendation that the costs of social obligations be borne by

Telecom alone (echoing the earlier Court of Appeal ruling). Such incon-

sistency suggests pragmatic political interest trading, rather than princi-

pled economic analysis.

8.3.3 Introduction of Industry- Specifi c Regulation with Effi ciency

Objectives

The government chose to give legislative force to almost all of the

Commission of Inquiry’s recommendations, introducing them into the

terms of the Telecommunications Act of 2001. The Act compromised

the pursuit of effi ciency objectives, to some extent, by pursuing other,

largely distributional, objectives. It established a Telecommunications

Commissioner within the Commerce Commission, and mandated TSLRIC

pricing for ‘designated services’ (‘retail- minus pricing’ was imposed for

lower- level ‘specifi ed services’). This pricing approach in part addressed

the distributional objectives of Telecom’s competitors, despite compro-

mises to dynamic effi ciency (Quigley, 2004).

However, the Act did include some pro- effi ciency measures. For

example, it ring- fenced the costs of the ‘Kiwi Share’ social obligations

as the Telecommunications Service Obligation (TSO), a charge the

Commissioner could levy on all industry participants. This helped to avoid

some of the worst dynamic effi ciency implications that would have arisen


had the Court of Appeal decision on the Clear case not been overturned

by the Privy Council. Furthermore, the Act created in Section 18(2) a spe-

cifi c obligation for the Commissioner to take account of effi ciency when

making decisions and recommendations:

in determining whether or not, or the extent to which, any act or omission will result, or will be likely to result, in competition in telecommunications markets for the long- term benefi t of end- users of telecommunications services within New Zealand, the effi ciencies that will result, or will be likely to result, from that act or omission must be considered.

The Act also (potentially) safeguarded the Commissioner’s independence

from political infl uence by including the Offi ce of the Telecommunications

Commissioner within the Commerce Commission. Under the Crown

Entities Act, the Commerce Commission is an Independent Crown Entity

(ICE), and thus has the least obligation of any of the Crown Entities

(statutory bodies outside of core government departments) to political

imperatives. An ICE must take account of government directives only

where it is specifi cally required to do so as part of its legislated duties (only

the judiciary has greater independence from political direction). In the

Commissioner’s case, taking effi ciency into account was a legislated duty

under the Telecommunications Act.

This statutory obligation to consider effi ciency suggests that, despite

some of the more radical Commission of Inquiry recommendations, on

balance the government believed that a balanced, total effi ciency objec-

tive remained important for the telecommunications sector, albeit slightly

diluted by the other provisions of the Act. Combined with the appar-

ent independence of the Commission, it could be interpreted that the

Telecommunications Act attempted to create a regulatory body, charged

with pursuing effi ciency, that was independent from risk of capture by

vested (and especially political) interests. At the time, this was seen as an

‘enlightened’ form of industry- specifi c regulation.

8.3.4 Testing of the Regime with LLU Decision (Dynamic Effi ciency

Prevails)

The enforceability and political acceptability of decisions made using the

legislated effi ciency mandate were explicitly tested when the Commission

undertook its statutory (Section 64) review into local loop unbundling

(LLU) in 2003. The investigation was noteworthy internationally for

employing a cost–benefi t analysis based upon a total welfare decision crite-

rion – the sum of consumer and producer surplus as opposed to consumer

welfare alone (Hausman and Sidak, 2005) – albeit subject to criticisms


regarding the treatment of investment in the model. The Commerce

Commission (under the mandate of the Offi ce of the Telecommunications

Commissioner) explicitly applied dynamic effi ciency principles in making

the ultimate recommendation not to proceed with unbundling.

The Commerce Commission found that: ‘The overall benefi ts from

unbundling are not suffi ciently persuasive to satisfy the Commission that

a regulated solution is warranted’. It considered platform competition (for

example, from wireless networks) likely to evolve and reduce the extent of

Telecom’s control of the bottleneck to access (Commerce Commission,

2003, para. 788). The Commission noted that, internationally, the experi-

ence of LLU was mixed in respect of increasing broadband penetration

(para. 792). The Commission also cited high costs of mandatory unbun-

dling, and made the critical point that the incentives for Telecom to invest

would be substantially reduced under LLU, and that this would have very

signifi cant eff ects upon the potential welfare gains for consumers (para.

794).

In place of LLU, the Commission recommended accepting an off er by

Telecom to make available a limited bitstream service. Given Telecom’s

imminent investment in a next generation network (NGN), a lesser form of

unbundled access would enable a limited amount of service- diff erentiated

entry, whilst preserving Telecom’s investment incentives and limiting

entrants’ exposure to stranded assets in the event of the NGN resulting in

the bypass of exchanges containing entrants’ equipment (Gans and King,

2004; Covec, 2004; Howell, 2007).

The Minister of Communications accepted the Commission’s recom-

mendations, apparently endorsing both the methodology used and the con-

clusions reached. The decision was, however, received with considerable

dismay by Telecom’s competitors and other interested parties, who would

undoubtedly have preferred a diff erent outcome. The LLU decision illus-

trated to industry participants that industry- specifi c regulation constrained

by an explicit effi ciency obligation could result in an economically rational

outcome, rather than bowing to pressures to meet specifi c competition

or redistribution agendas. However, the decision also showed that, in the

absence of public opportunities to deviate the regulator from legislatively

mandated effi ciency- based decision- making, lobbying eff orts would need to

concentrate on (less well- informed) decision- makers in the political arena.

8.3.5 Testing of the Regime with Mobile Termination Decision

(Competition Trumps)

The ongoing supremacy of the total effi ciency criterion and the balance

between its static and dynamic components as interpreted in the LLU


inquiry was further tested by the Commerce Commission’s inquiry into

mobile termination between 2004 and 2006. The initial inquiry was insti-

gated by the Telecommunications Commissioner, who explained that ‘fea-

tures of the mobile termination market that give rise to concerns about the

exercise of market power by mobile carriers’ had led to ‘complaints that

lack of competition in the mobile termination market means charges for

fi xed- to- mobile calls in New Zealand are unreasonably high’.

Once again, dynamic effi ciency considerations underpinned the

Commerce Commission’s cost–benefi t analysis, which led to a recom-

mendation in June 2005 that mobile termination charges for voice calls on

2G (second generation) networks, but not 3G networks, be regulated. The

recommendation stated:

the Act does not direct the Commission as to the weight that it should give to effi ciencies, as opposed to other considerations. This is a matter for the Commission to consider. Where there are tensions between short- term alloca-tive effi ciency and long- term dynamic effi ciency, the Commission takes the view that giving greater weight to the latter will generally better promote competi-tion for the long- term benefi t of end- users. (para. 28).

Notwithstanding, the Commission gave greater weight to distributional

considerations in this recommendation than in the LLU case, with con-

sumer welfare, rather than total welfare, providing the decisive criterion.

In August 2005, however, the Minister rejected the Commerce

Commission’s recommendation and ordered that a second review be

undertaken, suggesting that a change in political priorities had occurred

since the 2003 LLU inquiry. The second review was to reconsider ‘defi ni-

tional and implementation issues concerning 2G and 3G’ and to take into

consideration ‘commercial off ers made by Telecom and Vodafone follow-

ing the Commission’s fi nal report’. The ministerial redirection appears

to confi rm that lobbying of politicians was now the preferred method of

infl uencing regulatory decisions.

The Commerce Commission’s second report, in April 2006, recom-

mended that all fi xed- to- mobile voice calls on all technology types be

subject to regulation, on the basis that ‘substantial net benefi ts to end users

were likely to arise from making mobile termination a designated access

service’ (para. 32). The Commission defended using the redistributive con-

sumer welfare decision criterion, explaining: ‘where wealth transfers which

are sustainable and not themselves conducive to ineffi ciency are likely to

result from a measure promoting competition, the Commission ought to

give weight to such transfers in the cost–benefi t analysis’ (para. 34). The

Commission justifi ed including 3G technologies in its regulatory prescrip-

tion by noting that deployment had advanced considerably between the


fi rst and second decisions, to the extent that the Commission considered

existing 3G investments to be irreversible.

More surprising than this shift in recommended regulatory scope,

though, was the Commission’s explicit rejection of the supremacy of long-

term dynamic effi ciency over short- term competition- based static effi cien-

cies and distributive considerations in its decision- making criteria – a

substantial change from the criteria emphasized in the fi rst report.

The Commission argued that its statutory authority actually prioritized

the pursuit of competition (not effi ciency) as the prevailing sector objective.

Specifi cally, the Commission claimed that the Telecommunications Act

created a distinction between two provisions of the Commerce Act: Sections

36 and 47 on the one hand, which seek to promote competition by restrict-

ing the aggregation of market power and controlling its use, and Part IV on

the other hand, which focuses on the regulation of existing market power.

The Commission interpreted Part IV as requiring a focus on the net benefi t

to acquirers – that is, the Commission should take into account ‘the wealth

transfer that occurs in reducing the excessive profi ts of the regulated party’

(para. 46) – an apparent acknowledgement of static effi ciency and redis-

tribution as the primary purpose of regulatory activity. The Commission

deemed the Telecommunications Act to derive from the principles of Part

IV of the Commerce Act, and thus was in no doubt that in addressing the

tension between the promotion of competition (the means) and the pursuit

of total, long- run effi ciency (the end), the Telecommunications Act gave

primacy to competition, stating: ‘where there is a tension between the

net public benefi ts and promotion of competition, the statutory context

indicates that the primary consideration is the promotion of competition’

(para. 47). A paragraph later: ‘the Telecommunications Act is focused on

regulating access to promote competition. It does not provide a mechanism

that specifi cally allows for effi ciency considerations to take precedence

over the promotion of competition. Nor is there anything in the statutory

scheme to suggest that this should be the case’.

That is, diametrically opposed to the LLU decision criteria, the static

and dynamic effi ciencies arising from increased competition on exist-

ing networks in the short run were deemed to take precedence over the

longer- run effi ciencies arising from investment in new network technolo-

gies in Commission decision- making. This suggests that, even if it could

be demonstrated that restricting competition would result in greater total

effi ciency in the long run, and even ultimately the long- term benefi t of

consumers, the Commission must as its statutory duty promote short- run

competition. The conclusion is that in this decision, the Commission is

confi rming a statutory obligation to place the short- term interests of com-

petitors above even those of consumers.


This conclusion, therefore, begs the question of the purpose of Section

18(2) of the Telecommunications Act. For pursuit of competition to over-

ride pursuit of effi ciency, ‘consider’ must constitute only an obligation to

demonstrate that as some part of the decision process, effi ciency issues

were raised. The Commission evidently now deemed Section 18(2) to

contain no statutory obligation that effi ciency considerations should have

any material eff ect upon the fi nal recommendation if, in taking them into

account, the pursuit of competition was impaired. If the Commission’s

2006 view is accepted as legitimate, then it constitutes an admission that all

of the previous Commission decisions prioritizing effi ciency and weighing

long- term dynamic considerations over short- term ones were based upon

an erroneous interpretation of the government’s intentions when passing

the Act. In the historical context, this view is perplexing, given the appar-

ent acceptance of all previous effi ciency- based recommendations.

A plausible alternative explanation for the Commission’s radical depar-

ture from previous decision- making precedents is that the prevailing

political objectives changed between the 2003 LLU decision and the

2006 mobile termination decision. The ‘independent’ regulator had been

perhaps too successful in prioritizing long- run total effi ciency for political

(and notably electoral) preferences which prioritize delivery of short- run

changes. As an agent of political principals, the Commission became

subject to pressures to resile from its previous prioritization of total, long-

run effi ciency in favor of a set of more politically acceptable short- run

criteria. Evidence in support of this view includes the Governor General’s

(2005) ‘Speech from the Throne’ signaling that the incoming government

would be ‘advancing policies to ensure that the telecommunications sector

becomes more competitive and that we achieve faster broadband uptake

in line with our competitors’ and the fact that the Minister, when rejecting

the Commission’s fi rst mobile termination recommendation, instructed

the Commission to take account of alternative off ers made to the Minister

by the fi rms facing regulation. An acknowledgement of the Commission’s

subservience to ministerial control is contained in its acknowledgement

that ultimately ‘the role of the Commission is to recommend and it is for

the Minister to determine’ (para. 53).

The above observations on the Commission’s shift in decision- making

processes suggest that competitor lobbying of politicians for competitive

and redistributive interests to prevail successfully led not just to the change

of objective but also to a fundamental change in the locus of control

of New Zealand’s regulatory decision- making. Ultimately, it was the

Minister and not the Commissioner who began to exercise actual regula-

tory control. As further evidence of this shift, the Minister also rejected

the Commission’s second recommendation to regulate mobile termination


rates, instead favoring a set of undertakings by the potentially regulated

fi rms brokered by the Minister of Economic Development (the Minister of

Communications having declared a confl ict of interest as a consequence of

a legal dispute with one of the companies facing regulation).

8.3.6 Ministry- Led ‘Stocktake’ in Interest of Increasing Competition

The pursuit of competition in the telecommunications sector became an

increasingly important political issue as the incoming government began

implementing its political agenda. In December 2005 the Ministry of

Economic Development (MED) began a ‘stocktake’ of the telecommu-

nications industry, with its primary focus being ‘the broadband market

and our broadband performance as a factor in economic performance’

(MED, 2006). Given that the expertise to undertake the investigation lay

principally in the Offi ce of the Telecommunications Commissioner in the

Commerce Commission, the use of a ministry to undertake an assess-

ment of industry performance implies a lack of political confi dence in

the Commission and its processes. It also suggests a lack of certainty that

Commission- led analysis ‘taking account of’ effi ciency would deliver a set

of recommendations consistent with the government’s explicit competi-

tion agenda to which it was already committed. Unlike the Commission,

the Ministry was not explicitly bound by a requirement to take account of

effi ciency issues in its analyses.

The stocktake report released by the MED in May 2006 is notable for its

lack of principled economic analysis (see Howell, 2006 for a detailed dis-

cussion). In sharp contrast to the Commission’s analyses, no cost–benefi t

analysis was undertaken. The Regulatory Impact Statement accompa-

nying the stocktake proposals rejects the validity of the Commission’s

analysis in 2003, but specifi cally states that no new analysis was necessary

to quantify the eff ects of the stocktake’s recommendations. Rather, the

stocktake justifi ed full local loop unbundling and functional separation of

Telecom primarily on the basis that it believed the broadband market had

failed to meet a desired level of competitiveness (defi ned, arbitrarily, by

the share of connections sold by competitors to Telecom), and that invest-

ment by Telecom had been ‘insuffi cient’ (on the basis of slippage from an

investment schedule proposed in 2003 and an unscientifi c benchmarking

against a handful of other OECD countries).

Compounding the severe analytical shortcomings of the stocktake,

the process for reviewing its recommendations also lacked rigor and

transparency. Submissions on the stocktake’s proposals were heard by a

select committee comprised of generalist politicians rather than a panel of

expert commerce commissioners, and were not subject to the three- stage


Commission processes of a draft report, a quasi- court conference where

the recommendations and written submissions on the draft report by all

interested parties can be tested in a contestable manner, and a fi nal report.

Furthermore, the process was not subject to appeal or review on either

process or substance.

8.3.7 Changes to the Telecommunications Act Increase Political Infl uence

Thus, with little alteration and considerable political support, the stock-

take’s recommendations were enacted in December 2006. Along with the

provisions for full LLU and separation, a new section (19A) was added to

the Telecommunications Act requiring the Commission to take account

of any economic policies of the government that are communicated by

the Minister in writing. This appears to place the requirement to take

account of government economic policies on the same footing as the

requirement to take account of effi ciency in making recommendations as

per Section 18(2). If treated as an imperative, as the effi ciency obligation

was from 2001 to 2005, it would appear to explicitly to compromise the

Commission’s independence from the government by making it an instru-

ment via which government economic policy is enacted. Alternatively,

if interpreted in the light of the second mobile termination inquiry, the

new requirement could be subjugated, but only in favor of the pursuit of

competition – which, incidentally, was an economic policy of the govern-

ment that incorporated the requirement in the fi rst place.

Rather than clarify the Commission’s objectives, it appears that the

new obligation compounds the problem of multiple priorities for the

Commission, and reduces the possibility that any of the objectives will

be satisfactorily delivered. Moreover, it creates a further tension if, in

the future, government economic policies change again in a manner

that brings them into confl ict even with the competition imperative (for

example, nationalization of assets).

Together, the 2006 stocktake and Telecommunications Act amend-

ments signal that the New Zealand regulatory process no longer made

any pretence of being independent. The Ministry and politicians were now

fi rmly established, by a dint of formal legislative power, as the third party

alongside the courts and the Commission in governing industry activities.

In further confi rmation of the strong government role in sector regulation,

in May 2007 the Minister of Communications announced that he, rather

than the Telecommunications Commissioner, would be overseeing a new

process to functionally separate Telecom. The Minister (endorsed by the

cabinet) justifi ed his taking control of this process on the basis of ‘the

urgency attached by the government to the need to secure a clear outcome


on this matter in the shortest possible timeframe. Because this is a major

structural issue and not a matter of micro regulation, this was felt and is

still felt to be the appropriate way forward’ (New Zealand Government,

2007).

This declaration clearly indicated the government’s view that the struc-

ture of the telecommunications industry, where almost all investment

comes from the private sector, was the prerogative of political control,

whilst the role of regulators was a micro one of carrying out politi-

cally determined policies. Ministerial supervision of Telecom’s functional

separation marked the third time since the 2005 election that the formal

political power of the principal in the regulatory agreement was exerted in

order to take back control of duties ordinarily in the ambit of a regulatory

authority.

Such a view of political primacy is at considerable variance with

the prevailing views espoused by the OECD and the International

Telecommunication Union (ITU) that industry- specifi c regulation be

insulated from political processes to the greatest extent possible pre-

cisely to guard against the pursuit of short- term self- interested agendas.

However, under the present arrangements for sector governance in New

Zealand, it is highly unlikely that future regulatory decisions will be driven

primarily by economic effi ciency imperatives, or even requires that an

analysis taking account of the sectors’ specifi c economic circumstances

(for example, small- scale, high and sunk investment costs) be undertaken.

Rather, recent history suggests that returns to political lobbying will more

likely bear fruit (for both the incumbent and entrants) than economic

debate with the regulator.

8.4 CONCLUSION

New Zealand telecommunications regulation from 1987 to 2009 off ers

a cogent example of the vulnerability of even those regulatory agencies

designed with economic effi ciency objectives in mind to capture by politi-

cal principals in order to pursue alternative objectives. Over the course of

less than two decades, the pursuit of competition and the achievement of

short- term effi ciency gains as a political priority prevailed over the pursuit

of total effi ciency in the long run – even though initially, total, long- run

effi ciency was explicitly articulated as the prevailing political objective.

Reliance upon competition law alone was the starting point for the allo-

cation of industry regulatory governance to achieve the stated objective.

This was replaced by an industry- specifi c regulator, with a legislative

mandate to consider total effi ciency. However, over time this body has


been politically infl uenced towards the pursuit of competition and politi-

cally motivated subjective short- term redistribution issues and notably

competitor welfare, and its legislative mandate now requires that it con-

sider political objectives in carrying out this function.

Whereas, typically, political capture of regulatory processes occurs

informally as covert exertion of pressure and infl uence, the New Zealand

case illustrates not only informal pressure, but also overt political action.

Political dissatisfaction with both court adjudications and regulatory

recommendations based upon total effi ciency criteria has resulted in a

‘third body’ – the offi ce of the Minister of Communications and the affi li-

ated MED – assuming responsibility for a range of functions normally

delegated to the purview of an industry- specifi c regulator. In the absence

of legislative or constitutional formalization of this action, it is not clear

whether the result is intended as competition for, or a complement to, com-

petition law and regulation. What remains clear, however, is that current

political decision- making has given very little weight to any effi ciency

considerations, has reduced industry certainty as it lacks precedents, and

is subject to few of the checks and balances of either competition law or

regulatory processes.

New Zealand’s unique institutional structure appears to have evolved

because the explicit attempts by governments in the past to create arrange-

ments that prioritized the pursuit of total effi ciency above other considera-

tions, in a manner that refl ects the competitive reality of the small New

Zealand economy, have ultimately failed to withstand the threat of politi-

cal capture. The government initially relied upon competition law alone

to regulate the telecommunications sector, and subsequently, when intro-

ducing an industry- specifi c regulator, explicitly included effi ciency objec-

tives in the mandate given to the industry- specifi c regulator. However,

subsequent governments diluted the force of the effi ciency mandate, and

ultimately have captured the regulatory process as a means of furthering a

political agenda most likely driven by redistributive pressures. The erosion

of the effi ciency mandate began gradually, but has accelerated since 2005.

The chronology of erosion and capture illustrates the thesis that, despite

the best of intentions, ultimately governance arrangements in the telecom-

munications sector are determined by those with the political power to

make laws allocating responsibility for various tasks. Pursuit of effi ciency

is rationally justifi able from an economic perspective, and has been dem-

onstrated to be the objective of (if not perfectly achievable via) competi-

tion law. However, making effi ciency an explicit regulatory objective is

not sustainable in the long run. An objective regulator would not heed the

petitions of those seeking to capture the process for their own purposes,

shifting pressure onto the lawmakers in government. The government will


permit a continuing focus on effi ciency only as long as this is congruent

with its political objectives. But governments change and their objectives

change with them. Eventually, in the absence of the ability to exercise

informal control to capture the process, a successor government (which

cannot be bound by its predecessors) may reverse earlier decisions, either

changing the rules or taking over the regulatory task itself. The only reason

that the courts administering competition law can avoid such capture is

because their constitutional origins aff ord them more independence than

agencies that derive their mandate from political processes.

Given the lack of ability to enforce an effi ciency objective in the long

run via industry- specifi c regulation, the only sustainable means of doing

so would appear to be via competition law. Imperfect though competi-

tion law may be for the telecommunications sector, the New Zealand case

suggests that, in the absence of constitutional protections for a regulatory

agency also charged with the pursuit of effi ciency, it may be the only sus-

tainable institutional compromise.

NOTE

1. This chapter, prepared with editorial assistance from Seini O’Connor, is based on a working paper for the New Zealand Institute for the Study of Competition and Regulation (‘The end or the means? The pursuit of competition in regulated telecom-munications markets’ see www.iscr.co.nz).

REFERENCES

Alleman, J. and P. Rappoport (2005). Regulatory failure: time for a new para-digm. Communications and Strategies, 60, 105–21.

Armstrong, M. (2001). Access pricing, bypass and universal service. American Economic Review, 91, 297–301.

Arnold, T., D. Boles de Boer and L. Evans (2003). The structure of industry in New Zealand: its implications for competition law. In M. Berry and L. Evans (eds), Competition Law at the Turn of the Century (pp. 11–18). Wellington, New Zealand: Victoria University Press.

Blanchard, C. (1994a). Telecommunications regulation in New Zealand: the Court of Appeal’s decision in Clear Communications v Telecom Corporation. Telecommunications Policy, 18(9), 725–33.

Blanchard, C. (1994b). Telecommunications regulation in New Zealand: how eff ective is ‘light- handed’ regulation? Telecommunications Policy, 18(2), 154–64.

Blanchard, C. (1995). Telecommunications regulation in New Zealand: light- handed regulation and the Privy Council’s judgement. Telecommunications Policy, 19(6), 456–75.

Commerce Commission (2003). Telecommunications Act 2001 Section 64 review


and schedule 3 investigation into unbundling the local loop network and the fi xed public data network Final Report. Wellington, New Zealand. Available from www.comcom.govt.nz.

Connolly, S., and A. Munro (1999). Economics of the Public Sector. Harlow: Pearson Education.

Covec (2004). Benchmarking Telecom’s UPC service. Retrieved from www.comcom.govt. nz.

Economides, N. and L. White (1995). Access and interconnection pricing? How effi cient is the ‘effi cient component pricing rule’? Antitrust Bulletin, 40(3), 557–79.

Evans, L., A. Grimes, B. Wilkinson and D. Teece (1996). Economic reforms in New Zealand 1984–1995: the pursuit of effi ciency. Journal of Economic Literature, 34(December), 1856–1902.

Feldstein, M. (1972a). Distributional equity and the optimal structure of public pricing. American Economic Review, 62, 32–6.

Feldstein, M. (1972b). Equity and effi ciency in public pricing. Quarterly Journal of Economics, 86, 175–87.

Gans, J. and S. King (2004). Access holidays and the timing of infrastructure investment. Economic Record, 80(248), 89–100.

Governor- General Dame Silvia Cartwright (2005). Speech from the Throne. Delivered to Members of the House of Representatives at the State Opening of Parliament on 8 November. Transcript available at www.beehive.govt.nz/node/24330.

Hausman, J. and G. Sidak (2005). Did mandatory unbundling achieve its purpose? Empirical evidence from fi ve countries. Journal of Competition Law and Economics, 1(1), 173–245.

Howell, B. (2003). Building best practice broadband: bringing supply and demand together. ISCR Working Paper. Available at www.iscr.org.nz.

Howell, B. (2006). Submission: Telecommunications Amendment Bill. Presentation to the New Zealand Finance and Expenditure Select Committee. Available at www.iscr.org.nz.

Howell, B. (2007). A pendulous progression: New Zealand’s telecommunications regulation 1987–2007. ISCR Working Paper. Available at www.iscr.org.nz.

Kahn, A. (1970). The Economics of Regulation, Vol. I. New York: Wiley & Sons.Kahn, A. (1975). The Economics of Regulation, Vol. II. New York: Wiley &

Sons.Ministry of Commerce and The Treasury (1995). Regulation of access to

vertically- integrated natural monopolies. Wellington, New Zealand: Ministry of Commerce, The Treasury. Available at www.med.govt.nz.

Ministry of Economic Development (MED) (2000). Ministerial inquiry into the telecommunications industry: fi nal report. Wellington, New Zealand: Ministry of Economic Development. Available at www.med.govt.nz/upload/30006/fi nal.pdf, accessed 19 January 2010.

Ministry of Economic Development (MED) (2006). Promoting competition in the market for broadband services. Wellington, New Zealand: Ministry of Economic Development. Available at www.med.govt.nz.

New Zealand Government (2007). Delivering on telecommunications. Speech notes for address to TUANZ Telecommunications Day. Michael Fowler Centre, Wellington, New Zealand, 31 May 2007, http://www.beehive.govt.nz/node/29595 (offi cial website of the New Zealand Government).


Peltzman, S. (1976). Toward a more general theory of regulation. Journal of Law and Economics, 19, 211–40.

Quigley, N. (2004). Dynamic competition in telecommunication: implications for regulatory policy. Toronto: C.D. Howe Institute Commentary, February.

Schmalansee, R. (1981). The Control of Natural Monopolies. Lexington, MA: Lexington Books.

188

9. International regulatory comparisons: the evolution of IP- based fi ber

Scott Marcus and Dieter Elixmann

9.1 INTRODUCTION

Technological and market forces are driving network operators and elec-

tronic communication service providers throughout the world to migrate

their networks to Next Generation Networks (NGNs) based on the

Internet Protocol (IP).1 NGN access to the fi xed network is in the process

of being enhanced over time in many countries to provide higher speed

using fi ber- based technology. At an abstract level, one might imagine that

a change in underlying technology would have little impact on regulation;2

however, the evolution of the access network to IP- based fi ber implies sub-

stantial challenges for regulators. Various regulators in various countries

are fi nding somewhat diff erent solutions to these challenges.

Two broad families of technical approaches exist. One approach is

known as fi ber to the home (FTTH) or fi ber to the building (FTTB). The

other is very high speed DSL (VDSL), which is associated with fi ber to the

curb/cabinet (FTTC) (because fi ber is built out to the street cabinet but

not all the way to the individual home or building). Whether the migration

is to FTTC/VDSL or to FTTB/FTTH, traditional solutions to incumbent

market power typically become diffi cult to apply. For FTTC/VDSL, the

natural point of interconnection (PoI) for purposes of network access

moves from the main distribution frame (MDF) to the far more numerous

street cabinets; however, access to street cabinets is potentially diffi cult

and costly, calling into question the practicality of local loop unbundling

(LLU) as a competitive remedy. For FTTB/FTTH deployments to mul-

tiple dwelling units, many challenges exist regarding building wiring. As

an additional problem for point- to- multipoint FTTB/FTTH, there are

uncertainties as to the practicality of unbundling passive optical network

(PON) solutions. All of these challenges call into question the practicality

of LLU as a pro- competitive regulatory remedy.3

International regulatory comparisons 189

Various countries have considered these issues, but the pace of regula-

tory proceedings has been conditioned in each case by developments and

market evolution in that country. In both the Netherlands and Germany,

the incumbents (KPN and DTAG) are seeking to move rapidly to replace

the traditional fi xed access network with a VDSL- capable network. In

France, the incumbent as well as its competitors are moving quickly to

deploy FTTx in a number of major metropolitan areas. In Japan, fi ber

deployment has been very successful, with many of the deployments

undertaken by third parties other than the incumbents (NTT East and

NTT West). In the US, incumbent operators AT&T and Verizon have

made substantial investments in fi ber for their respective customers. Our

empirical analysis refl ects the state of the world as of the fi rst half of

2008.

Section 9.2 discusses the regulatory challenges relevant to FTTC/VDSL

and FTTB/FTTH solutions. Section 9.3 discusses regulatory solutions

that have been attempted in various countries. Section 9.4 provides a com-

parative assessment, in terms of the degree to which competition has been

preserved, and the extent of fi ber- based deployment and adoption. Section

9.5 provides concluding remarks and a few recommendations.

9.2 REGULATORY CHALLENGES

This section explores the various regulatory challenges associated with

migration to IP fi ber- based Next Generation Access. Section 9.2.1 explores

challenges in an FTTC/VDSL context. Section 9.2.3 discusses problems

that are unique to an FTTB/FTTH environment. Finally, section 9.2.3

reviews problems that are common to both. Our focus in this chapter is on

challenges at the access level of the network; consequently, we disregard a

broad array of other regulatory concerns raised by the migration to NGN

access from copper to fi ber, notably including possible interactions with

the interconnection regime (which has more to do with the NGN core than

with the access network).

9.2.1 FTTC/VDSL

In this section, we consider the challenges the migration to FTTC/VDSL

poses to pro- competitive regulation. First, we consider challenges in

general; then we review specifi c challenges to the business model that have

been studied in the literature. Finally, we consider the degree to which bit-

stream access, which is largely unperturbed by the migration from ADSL

to FTTC/VDSL, could substitute for both LLU and shared access.


The deployment of FTTC/VDSL poses challenges to local loop unbundling

(LLU)

The fi rst- mover for FTTC/VDSL deployments has often been the incum-

bent. Examples include Belgium (Belgacom),4 Germany (Deutsche

Telekom), Italy (Telecom Italia)5 and the Netherlands (KPN).6 With an

(all- IP) FTTC/VDSL strategy, fi ber is deployed throughout the access

network up to the street cabinet. Wherever fi ber infrastructure is deployed,

copper infrastructure is removed (or taken out of service). VDSL access

devices (DSLAMs – DSL access multiplexers) are generally deployed at

the street cabinet, in order to be close enough to the end- user to achieve

suitably high bandwidth; consequently, the main distribution frames

(MDFs) typically no longer serve as active network nodes. This means, as

previously noted, that the point of interconnection (PoI) moves from the

MDF to the street cabinet. This shift implies multiple challenges (in com-

parison to copper- based ADSL – Asymmetrical Digital Subscriber Line)

for would- be competitors who seek to replicate the incumbent’s FTTC/

VDSL infrastructure:

There are far more nodes to connect to, because there are far more ●

street cabinets than MDFs.

There are far fewer end- users per street cabinet. ●

There are likely to be limitations with the physical characteristics of ●

the street cabinets, including (alleged) lack of space, heat dissipa-

tion, and so on.

There is no realistic possibility to provide separate, additional street ●

cabinets for competitors due to the resistance of city governments

and home owners (whose premises are potential locations for a

separate street cabinet).

There are likely to be high barriers or expenses in connecting to the ●

existing street cabinets, because the competitor is unlikely to have a

cost- eff ective alternative to the duct infrastructure that the incum-

bent has inherited from monopoly times.

Thus, incumbent FTTC/VDSL deployments are likely to pose serious

challenges for existing wholesale services based on local loop unbundling

(LLU), including shared access models.

Sub- loop unbundling and the business case for FTTC/VDSL

Three major studies have found that the business case for sub- loop unbun-

dling (SLU) is likely to be problematic.7 A study for the Dutch regulator

OPTA (Analysys, 2007a) found that the use of sub- loop unbundling (SLU)

by an alternative provider would be economically viable as an alternative


to continuing to use LLU only under very limited conditions. Some

combination of: (1) high market penetration; (2) higher average revenue

per user (ARPU) than has typically been achieved to date in Europe; or

(3) a substantial reduction in wholesale prices charged by the incumbent

would be required. These requirements are collectively and individually

unlikely to be fulfi lled. For example, a mass- market competitor could be

viable with both: (1) a market share greater than 55 percent of all broad-

band lines (including cable) in areas served; and (2) an increase in ARPU

across all broadband users of 10 euros per month. Alternatively, an SLU

competitor with a more limited footprint could be economically viable

with: (1) a 10 percent market share of all broadband lines in areas served,

corresponding to about 1000 of the largest street cabinets in the densest

urban areas; in conjunction with (2) a 50 percent reduction in the intercon-

nect and wholesale tariff s from KPN for SLU line rental, co- location and

links to the street cabinets; and (3) an increase in ARPU of about 9 euros.

Again, these conditions are unlikely to be fulfi lled.

A second study that was conducted for the Irish national regulatory

authority ComReg (Analysys, 2007b) reached similar conclusions. SLU

is subject to much stronger economies of scale than is LLU. SLU is not

as commercially attractive as LLU, even when the competitor concen-

trates on large street cabinets (more than 300 lines). The largest costs

are the line rental charge, cost for the street cabinet and the backhaul

link to the MDF. These are the points to be addressed fi rst by regulatory

means.

A third study (J.P. Morgan, 2006) compared the incremental invest-

ments and costs associated with FTTC/VDSL to those of ADSL. Both

were assessed using a greenfi eld approach, which is to say that all invest-

ments are assumed to be new. Compared to an ADSL competitor, a

VDSL alternative operator incurs:

additional costs for unbundling at the street cabinet; ●

additional costs to deploy fi ber between the MDF and the street ●

cabinet;

additional costs of the backhaul network due to the greater capacity ●

requirements implied by VDSL access.

The overall incremental cost depends in part on the assumed network

topology, star or ring. In the worst case, an alternative operator with a

market share of 25 percent incurs additional costs of 11 euros per access

line per month. Like the two previous studies, J.P. Morgan (2006) shows

that FTTC/VDSL deployment is economically feasible for an alternative

operator only under very strong conditions. Even with a 40 percent market


share, the VDSL alternative operator’s profi t per subscriber decreases by 2

euros per month compared to the comparable ADSL case.

Bitstream access to FTTC/VDSL

Given that SLU is viable only under (very) limited conditions, bitstream

access must be considered as a practical alternative. The economics of

bitstream access for the incumbent will change somewhat with the migra-

tion from ADSL to FTTC/VDSL (due, for example, to changes in the

number of aggregation points), but the viability of bitstream overall does

not appear to be threatened. A troubling question for the regulator is

whether bitstream access alone is suffi cient. Europe has sought to maintain

a ladder of investment with four rungs: resale (which provides no service

diff erentiation), bitstream access, shared access, and full LLU. The move

to FTTC/VDSL apparently breaks two out of three rungs of the ladder. Is

bitstream alone enough to ensure eff ective competition?

9.2.2 FTTB/FTTH

This section reviews the business and regulatory challenges posed by

FTTB and FTTH and present an analysis of their implications for busi-

ness models.

The deployment of FTTB/FTTH poses diff erent challenges

One can fi nd numerous examples of fi ber to the building (FTTB) or to the

home (FTTH) in Asia (Hong Kong, Japan and Korea), Europe (France,

Sweden and Italy), and the United States. While FTTC/VDSL players are

usually incumbents, FTTB/FTTH players can be either incumbents or

competitors (or sometimes both, even in the same country).

The two common architectural options are (1) passive optical network,

or PON (which is a point- to- multipoint infrastructure containing no active

electronic equipment where a single fi ber path branches to several houses);

or point- to- point (which is comprised of a single dedicated fi ber path from

the provider’s network node to the consumer’s building, that is, between

the optical line terminal (OLT) and the optical network unit (ONU). Civil

engineering costs and other diffi culties (digging up the streets, getting the

permissions and rights of way) dominate the cost of deployment. Where

aerial deployment is permitted, however, or where existing ducts can be

used, the cost of running fi ber can be greatly reduced. Key regulatory chal-

lenges include:

In- house wiring: for multiple dwelling units (MDUs), running ●

fi ber to individual apartments implies rewiring the building. Few


buildings are suitable for more than one set of fi ber. This adds a new

dimension to the last mile problem.

Sharing of PON infrastructure: for point- to- multipoint PON ●

systems, it is unclear how to provide an LLU equivalent to competi-

tors. Sharing would appear to be expensive if not impossible.

The business case for FTTB/FTTH

The J.P. Morgan (2006) study found that the investment required to

pursue an FTTB strategy can be up to 12 times higher than those required

for traditional ADSL access, and up to 2.5 times higher than for VDSL

access. If, however, existing ducts can be used for the deployment of fi ber,

then the investment required to deploy FTTB is only slightly higher than

in the VDSL case. The generic greenfi eld business case developed by JP

Morgan refl ects the following baseline model assumptions, which from

our perspective seem to be plausible:

The model is calibrated on a 25 percent market share. ●

The operator has no access to duct infrastructure. ●

There is no ARPU or market share gain due to FTTB/FTTH. ●

For each FTTH customer acquired the operator saves 13 euros per ●

month regarding LLU and interconnection.

There are no changes regarding other operating expenditure (OPEX) ●

components due to the FTTH rollout.

The weighted average cost of capital (WACC) is equal to 8 percent, ●

the tax rate is equal to 35 percent and the depreciation period is

equal to 20 years.

Under these baseline model assumptions, the payback period for the

investments in FTTB/FTTH is 16 years. The net present value (NPV) of

the investments is nonetheless negative, at - 500 euros per subscriber. A

market share of just 40 percent generates a positive NPV. The investments

can, however, be economically feasible with a market share of 25 percent

if the investments can be reduced to a level of 2000 euros per subscriber,

if additional ARPU is possible, or if additional savings are possible.

Regarding the latter, very important profi tability eff ects arise with savings

due to the deployment of fi ber. Provided that 50 percent of the deployment

costs to and within the building can be saved (by utilizing existing ducts

and in- house sharing), then the investment costs can be reduced to a level

of 1500 euros per subscriber. In this case, the payback period decreases

from 16 years to ten, and the investments become profi table with an NPV

of more than 200 euros per subscriber. A similar result would be obtained

if the ARPU were increased by 6 euros per customer per month.


9.2.3 The FTTC/VDSL and FTTB/FTTH Deployments: some Common

Issues

This section reviews a range of issues common to both, the FTTC/VDSL

and FTTB/FTTH deployments. Particularly, it deals with the problems of

market defi nition, the issues of network topology and the challenges they

pose to the traditional regulatory models.

Market defi nition

The identifi cation of markets is fundamental to the European regula-

tory framework. Consistent with numerous communications from the

European Commission, we have treated the various forms of fi ber- based

IP access in this chapter as being part of a single continuous market that

also includes conventional broadband (such as ADSL and cable). It is

possible that applications that are heavily dependent on the bandwidth

that only fi ber can off er will become more prevalent over time. At that

point, it might be appropriate to treat fi ber- based access as a distinct

market segment from conventional broadband, inasmuch as conven-

tional broadband will no longer be an adequate substitute for fi ber- based

access.8

Change of network topology

The migration to Next Generation Access (NGA) will result in changes

in the number and location of points where competitors can get access

to the network of the incumbent (henceforth also called Points of

Interconnection, or PoIs). The phasing out of MDFs9 for competitive

access is a key point. More generally, IP- based NGNs will be fl atter than

today’s networks, and will therefore off er fewer points of interconnection

as well. These changes in network topology are likely to lead to stranded

investments (equipment that was deployed in the past, is not yet fully

depreciated, but is no longer needed) both for the incumbent and for the

alternative operator. KPN’s approach in the Netherlands suggests that the

incumbent can benefi t by selling off the real estate associated with facilities

that are no longer needed. These benefi ts can off set the cost of stranded

incumbent investments. Competitive operators are less likely to benefi t

from such changes.

Stranded investments are most likely to be important in the transition

period from switched network to IP- based NGN. They need to be refl ected

in the regulatory environment of the transition period. Transparency

regarding the network deployment plans of the incumbent is crucial. The

less transparency on the part of the incumbent, the greater will be the risk

of stranded investments on the part of competitors.


Changes in regulated costs

Numerous regulatory decisions rest on assumptions about the cost

of the network. The migration to fi ber- based NGN access will, at a

minimum, require regulators to upgrade their cost models. This migra-

tion potentially raises more complex issues than the implementation of

ADSL, because the investments in fi ber are so much larger, and also

because they are often associated with fundamental changes in the core

network.10 The overall migration to fi ber- based IP requires massive

investment at the outset, but presumably results in lower ongoing

operating costs when the transition is complete. At the same time, the

network as a whole may incur higher operating costs during a period

where the circuit- switched public switched telephone network (PSTN)

and the packet- switched NGN must operate in parallel. Ongoing oper-

ating costs should eventually decline, but initially they are likely to

increase. What is an appropriate regulated rate of return under these

circumstances?

Ofcom explained analogous issues in developing a weighted average

cost of capital (WACC) for British Telecom.11 The WACC provides a

measure of what might constitute a reasonable return on BT’s investment

of capital. Ofcom found it advantageous to disaggregate the WACC,

computing diff erent WACCs for diff erent parts of the business in order to

accommodate diff erent levels of risk (beta). In the end, it did not specifi -

cally address the risk associated with the migration to NGN access, but it

left open the door to possible future use of more innovative approaches,

including the application of REAL OPTIONS.

9.3 REGULATORY RESPONSES: COUNTRY- SPECIFIC APPROACHES

Diff erent regulatory approaches are visible in diff erent countries. Note

that the Netherlands and Germany are (chiefl y) characterized by FTTC/

VDSL roll- outs, while France and Japan are primarily characterized by

FTTB/FTTH roll- outs. In the US, both are present (with AT&T promi-

nently following a path of FTTC/VDSL, and Verizon a path of FTTB/

FTTH).

9.3.1 Netherlands

KPN (as of 2008), is about to deploy a nationwide FTTB/VDSL infra-

structure.12 OPTA, the regulator for the Netherlands, initially was open

to KPN’s all- IP deployment plan and planned to fi nd pertinent regulatory


rules to shape the migration period. However, the Analysys study (see

section 9.2.1) called into question the economic viability of sub- loop

unbundling from the street cabinet. Several rounds of discussion and

negotiations followed. It has now (as of mid- 2008) been concluded that the

current MDF related services will be maintained until the middle of 2010,

that there will be an improved off ering regarding high- quality wholesale

broadband access, and that a substantial portion of MDF locations will

remain usable for unbundled access. Moreover, OPTA has asked market

participants also to discuss alternatives like Ethernet access on fi ber and

interconnecting leased lines.

9.3.2 Germany

In February 2007, the European Commission launched an infringement

procedure against Germany in response to a specifi c clause in the new

German telecommunications law that opened up the possibility of a

regulatory holiday for DTAG’s FTTC/VDSL deployment (Marcus and

Elixmann, 2008). As a result of the ongoing market defi nition and market

analysis process, the BNetzA (the German regulator) imposed an obli-

gation on DTAG in June 2007 to open up the ducts between the MDF

and the street cabinet to competitors. In the event that access to ducts is

not possible (for example, due to technical reasons or limited capacity),

DTAG must off er competitors access to dark fi ber.

Competitors in Germany view this as a step in the right direction;

however, they claim that access to ducts (dark fi ber) is only one of the

elements that make up a viable business case. Equally important, they

argue, is unbundled access to fi ber and copper within the street cabinet.

Moreover, they are demanding (bundled) access to the hybrid local loop,

consisting of copper and fi ber, at the MDF. It is clear that there are still

substantial information asymmetries between DTAG and the competitors

as regards duct availability and space.

On 13 May 2008, the BNetzA issued a decision regarding the price of

the IP bitstream access that DTAG is obliged to off er to its competitors.

With IP bitstream access, DTAG sells a competitor a regulated wholesale

service whereby data is transmitted to and from the end- user’s access line

over DTAG’s concentrator network, and is handed off to the competitor

at a designated DTAG PoI. This is the fi rst time that IP bitstream has

been meaningfully available in Germany (although Germany has well-

established LLU rules for copper- based LLU). This eff ectively opens a

new rung on the ladder of investment, which may partly compensate for

the lack of access to fi ber- based IP access.


9.3.3 France

ARCEP, the French telecommunications regulator (Autorité de Régula-

tion des Communications Électroniques et des Postes), fi nalized two

consultations in September 2007, focusing on access to ducts, and sharing

of in- house infrastructure in an FTTB/FTTH environment. Regarding

access to ducts, ARCEP argued that competitors must have access to

FT’s infrastructure in order to establish fair competition in the high- speed

broadband market. Hence, the objective of regulation should be to set

appropriate incentives for investments in local loop infrastructure. A pos-

sible implication could be more symmetric regulation. Moreover, ARCEP

sees no need to impose functional separation as has been done in the UK.

ARCEP notes that it cannot be effi cient for each fi ber- based network

operator to deploy its own fi ber and optical connectors in each building

and each apartment. Moreover, residents do not benefi t if there is ineffi cient

and excessive deployment within their buildings and apartments. Finally,

ARCEP favors the principle that end- users have the opportunity to switch

their broadband network operator without the need to move from one

location to another. Thus, they see the need for infrastructure sharing.

In June 2007, Free/Iliad had sued France Telecom (FT) before the anti-

trust authority over FT’s allegedly anticompetitive behavior for refusing

to give its competitors access to its civil engineering infrastructures. The

Conseil de la Concurrence issued an important decision on 12 February

2008. It found that FT’s holding of civil engineering infrastructures is

likely to give the company a particular responsibility, notably including

not to distort the play of competition on the very high- speed budding

markets in keeping for itself the use of the infrastructures and refusing to

allow its competitors to use them, or giving them discriminatory access.

FT started to deploy optical fi ber in its civil engineering infrastructures

with the copper local loop, while postponing its response to competitors’

request for having access to the same infrastructures. The Conseil has not

implemented any immediate remedial measures, considering that there

was no serious or immediate infringement to the sector; nonetheless, the

Conseil decided to carry on with the investigation on the merits of the case.

Meanwhile, France Telecom made commitments to ARCEP to create an

operational off er for access to its infrastructure. The Conseil was generally

satisfi ed with the reactions of alternative operators to the fi rst tests.

9.3.4 Japan

Japan is the country with by far the highest FTTB/FTTH penetration in

the world. There were more than 11 million subscribers at the end of 2007.


The most popular broadband service in Japan is still ADSL. On the other

hand, there has been a remarkable shift to FTTH/FTTB over the past four

years (2006–09).13 The Next Generation Broadband Strategy 2010 formu-

lated in August 2006 by the Japanese Government aims at eliminating the

non- broadband areas (Taniwaki, 2008). By fi scal year 2010, broadband

service should be available to 100 percent of the population; and super

high- speed broadband (FTTH) should be available to 90 percent of the

population.

In Japan, there is infrastructure competition on the very last mile, at

least in the largest metropolitan areas. Important drivers of this kind of

infrastructure competition are the density of the population, the fact that

aerial deployment of fi ber on the very last mile is permitted in Japan (in

contrast to much of Europe), and regulation. Even so, NTT’s market

position with respect to fi ber deployment is very strong. As of the end of

March 2007, the combined FTTB/FTTH market share of NTT East and

West by number of lines was 78.9 percent, while the FTTB/FTTH share

by revenue was equal to 69.0 percent according to Taniwaki (2008).14 For

many years, the Japanese broadband market has been built around an

open access regime. Take- up of ADSL took off rapidly once collocation

and unbundling rules for the access networks of NTT East and West were

established in the middle of 2000. Although unbundled access continues

to be an obligation for fi ber going forward, these obligations have been

less eff ective than the previous copper- oriented regime; thus, the role of

unbundling going forward in Japan is unclear.

Japan is on the verge of changing its policy due the anticipated

transition to full IP- based networks. In October 2006, the Japanese

government launched a New Competition Promotion Program 2010.

Japanese policy appears to be moving from a world of ex ante regula-

tion to a world of ex post regulation. This shift appears to imply sub-

stantial easing of price and tariff regulations, and the introduction of a

competition review mechanism. Currently, the fi ber access of NTT East

and West is unbundled, and its charge is calculated on the basis of the

estimated actual cost of fi ber deployment from 2001 to 2007 (that is,

for seven years). On the basis of an exchange rate of 160 yen per euro,

Katagiri (2008) reports a current price of 31.7 euros per fi ber.15 Katagiri

points out that this price regime might be unfavorable for NTT’s

competitors. The reason is that FTTB/FTTH is usually provided via a

passive optical network architecture, and in NTT’s network one fi ber is

split to eight users. Given that the NTT’s market share is more than 70

percent (see above), the cost per user would be quite diff erent for NTT

in comparison to its competitors.


9.3.5 The United States

Regulation – or rather deregulation – of fi ber- based access in the United

States is relevant both to VDSL and to FTTB/FTTH deployments. The

United States during the years 2001–08 has consistently pursued a policy

of deregulation of last mile access, possibly in the hopes of stimulating

deployment of fi ber- based solutions. This refl ects a dramatic change from

prior US communications policy (Marcus, 2005).

The most signifi cant change in regard to fi ber- based access came in

2003 with the Federal Communication Commission’s (FCC) Triennial

review (FCC, 2003). This decision exempted fi ber from loop unbundling

obligations. In a series of related decisions, the FCC abolished non-

discrimination obligations for telephony incumbents in regard to IP- based

services; eliminated obligations to implement shared access (line sharing);

and exempted broadband services delivered over cable television from

pro- competitive regulatory remedies. It is important to remember that the

market in the United States diff ers from that of most other countries in

that cable operators provide a larger fraction of broadband access than do

telephone companies. This has established a signifi cantly diff erent market

dynamic than that of other countries.

9.4 COMPARATIVE ASSESSMENT OF THE REGULATORY APPROACHES

In this section we discuss Next Generation Access policy in the context

of a trade- off between investment and competition, and we consider the

degree to which regulatory approaches have been eff ective.

9.4.1 Is there a Trade- off between Investment and Competition?

What are the likely implications of regulation for competition, for deploy-

ment and adoption of fi ber- based access, and for consumer welfare? How

will regulation impact on the incentives and viability of business models

in the market? Does regulation negatively impact incentives to invest, for

incumbents and for competitors? How will it impact on the intensity of

infrastructure- based and service- based competition?

These questions necessarily involve complex trade- off s, the resolutions

of which are not clear. What is clear is that regulators are interested in

both competition and innovation, and that there is sometimes a tension

among these objectives. For Europeans, this tension is manifested in

Article 8 of the Framework Directive, which simultaneously calls on


regulators to ensure ‘that users . . . derive maximum benefi t in terms of

choice, price, and quality’, and ‘that there is no distortion or restriction of

competition in the electronic communications sector’, but at the same time

to ‘[encourage] effi cient investment in infrastructure, and [promote] inno-

vation’.16 What is the ‘poor’ regulator to do if these goals are somewhat at

odds with one another?

The arguments in favor of pro- competitive regulation are well estab-

lished in network industries. There are key assets that would not be

cost- eff ective for competitors to replicate. In the absence of regulatory

intervention, the industry would tend towards monopoly, resulting in a

loss of consumer choice, in a transfer of surplus from consumers to sup-

pliers, and typically in deadweight social loss ineffi ciencies. If suffi cient

infrastructure- based competition were present, it would be appropriate

to withdraw last mile regulation and to leave matters to the unregulated

market. Today, these conditions do not appear to be satisfi ed anywhere,

including, in our opinion, North America. Thus, our sense is that pro-

competitive interventions on last mile access will need to be maintained

for the foreseeable future, even if that were to imply some cost in the

speed of fi ber- based access deployment. It is by no means certain that

maintenance of regulatory remedies on last mile access has a dramatic

adverse eff ect on deployment. In the case of broadband, Europe has

achieved near parity in penetration with the United States, despite inten-

sive regulation, a dearth of cable television infrastructure, and a simply

enormous head start for the United States. A number of studies have

analyzed empirically the interrelationship of investing in broadband

infrastructure/broadband penetration and regulation. However, there

is no unique conclusion regarding the issue of whether willingness to

deploy is conditioned on lack of regulation (Friederiszick et al., 2007;

Wernick, 2007).

Having argued that it is appropriate in all or nearly all cases to main-

tain regulation on the last mile, we now come to the question of how it

should be implemented. The two primary alternatives would appear to be:

(1) imposition of conventional regulatory remedies as is practiced in the

European Union; or (2) functional separation, as has been implemented

in the UK by Ofcom and British Telecom (BT). The functional separa-

tion model is attractive, but it is still a bit soon to assess its true eff ective-

ness. Since BT has not implemented fi ber- based solutions yet, there is no

working example to point to (and indeed, one must ask whether func-

tional separation has somehow discouraged the deployment of fi ber in the

UK). These questions are too large to deal with in this chapter, but they

will unquestionably occupy the attention of experts in the years to come

(Kirsch and von Hirschhausen, 2008).


9.4.2 How Eff ective have Current Approaches been in Promoting

Investment and Competition?

This sections aims at illuminating how eff ective current regulatory

approaches have been in promoting investment and competition. To this

end, we focus on Germany, France, Japan, the Netherlands and the US.

We are addressing both the deployment and adoption of fi ber and the

competitive structure in these countries.

Germany

As of May 2008, DTAG has deployed FTTC/VDSL infrastructure in

27 cities, and ADSL 2+ infrastructure in about 750 cities, thus, reaching

around 17 million households. By the end of 2008, a further 13 cities will be

connected to the VDSL network, and ADSL 2+ will be available in about

1000 cities. Thus, high- speed DSL broadband access (fast enough for

Internet Protocol television – IPTV, for example) will be potentially avail-

able to 20 million households (about half of all households in Germany).

As of now, there is no publicly available information about phasing out of

MDFs. Conventional wisdom suggests that FTTC/VDSL is only a step on

the road to a long- term deployment of FTTB/FTTH.

Although its overall market share still is considerable,17 there are at

least two main driving forces for DTAG’s FTTB/FTTH deployment

activities. First, the speed of market losses regarding telephony access

lines is considerable: In the period 2006–2008, DTAG has lost about 2

million access lines per annum to competitors. Second, there is fi erce price

competition in Germany which has changed more and more to an overall

fl at rate regime for triple play services (VoIP telephony, broadband

access, Internet access). As a result, DTAG has had to respond to capital

market expectations in order to sell a convincing new growth story. In

this respect, however, DTAG has not been that successful. Its share price

has gone down from an all- time high of more than 100 Euros in 2000 to

a level on or below the fi rst IPO quotation price of around 14 Euros since

2005.

Regional competitors in Cologne, Munich, and Hamburg have launched

(or are planning to launch) FTTB/FTTH infrastructure. These ventures,

however, are concentrated on densely populated areas. An important

factor for their business case is savings from LLU wholesale services cur-

rently purchased from DTAG. Thus, a lower LLU price would have made

these investments of the competitors less likely. Cable modem access still

plays a minor role in the German broadband market, that is, intermo-

dal competition is still relatively low; however, the cable operators have

upgraded their networks in order to off er HFC (hybrid fi ber coaxial based)


triple play services. Even though cable represents a small fraction of the

installed base for broadband in Germany, it represents a quite substantial

fraction of the new customers added. As we have mentioned in section

9.3.2, the economic conditions for the IP bitstream access service off er-

ing of DTAG are now clear; however, it is far too early to assess how this

service is going to reshape the German competitive landscape.

France

In France, there are ambitious FTTB/FTTH deployment plans by the

incumbent and competitors alike (although geographically mainly focus-

ing on big cities with very dense population and a high percentage of

MDUs). A key driver for this development is the presence of man- high

sewage channels leading into many buildings, a circumstance that repre-

sents a special advantage for Paris and certain other French cities.

The regulator in France is very active and successful in convincing city

authorities to charge competition- friendly prices for providing access

to their infrastructure (for example, sewage channels). Policy- makers in

France are also monitoring the discussions with housing companies in

order to establish a win–win situation as regards access to in- house infra-

structure by fi ber deployers.

Japan

A far- reaching unbundling policy (for copper and more recently for fi ber)

has been a decisive factor for the rapid adoption and diff usion of xDSL.

Moreover, aerial deployment on the fi nal part of the loop brings costs

down. ADSL competition was fi erce in Japan. This, in turn, has led to

fi ber- based infrastructure competition, although the two incumbent com-

panies (NTT East, West) still account for an FTTB/FTTH market share

of more than two- thirds.

Netherlands

Unlike in other European countries, there is substantial publicly avail-

able information regarding the deployment plans of the incumbent in the

Netherlands. The regulator is in principle open to the next generation

fi ber- based deployment. The originally very ambitious deployment plan

by the incumbent (phasing out six out of seven MDFs by 2010) has been

postponed by at least 2–3 years due to regulatory decisions. Thus, there

will be no major impact on competitors in the short and medium term.

The United States

The withdrawal of pro- competitive regulation in the United States

has had more or less predictable eff ects. On the one hand, there is a


reasonably good rollout of fi ber- based access to the home, using a mix

of VDSL and of FTTH/FTTB technologies.18 On the other hand, there

has been the predictable negative eff ect on intramodal competition (that

is, competition among telephone companies for a given customers). For

traditional copper- based broadband, intramodal competition is negligi-

ble, having declined since the Triennial Review to some 3.1 percent of

ADSL lines. Intramodal competition over cable was never mandated,

and is negligible.19 For fi ber deployment, 91 percent has been under-

taken by incumbents large and small, and only 9 percent by competitors

(CLECs).20

Advocates for the US approach will argue that consumers have many

access options. They may, for instance, point to FCC statistics that show

robust deployments of wireless broadband. Our sense is that these argu-

ments are not well founded, for a number of reasons. The FCC counts as

broadband many services that nobody else would count as broadband,

including services that are less than 200 Kbps in the slower direction, and

mobile services where it is not clear that the subscriber is using the broad-

band and possibly does not view it as a substitute for a wired connection.

For the great majority of Americans, the realistic choices are broadband

from one cable operator or broadband (over copper or fi ber) from one

telephone company.

In assessing the overall eff ect of the FCC’s deregulatory approach, it

is diffi cult to say what the net eff ects have been. It seems clear that there

has been far less broadband deployed and adopted than one might have

expected given the long and early lead that the United States enjoyed,

and also the widespread availability of cable television. At the same time,

the robust rollout of fi ber is defi nitely an undeniable strength. How one

assesses the relative balance of the two necessarily refl ects complex judg-

ment calls: What is the true value to the consumer of the greater band-

width available with fi ber? What is the impact on the consumer of the lack

of intramodal competition? One must also bear in mind that the lack of

intramodal competition does not quite imply a re- monopolization of the

US market. Most Americans can choose cable broadband as an alter-

native to broadband from the telephone company – that is, two decent

options are available in most cases. It has led, not to monopoly, but rather

to an eff ective duopoly.

9.5 CONCLUDING REMARKS

The regulatory approaches in the countries that we considered can be

broadly grouped into three categories:21


Maintenance of pro- competitive remedies (France, Japan, ●

Netherlands).

Implementation of functional separation (UK). ●

Radical deregulation (the United States, and to some extent ●

Germany).

Can we say anything about the relative merits of the approaches? Which

approach is best? If there is anything that emerges clearly from compara-

tive assessment of diff erent regulatory approaches in diff erent countries, it

is that they started in diff erent places, and that they have achieved diff erent

results. The process of fi ber- based NGN access deployment can be viewed

as an enormous feedback loop, beginning with the intrinsic characteristics

of each country prior to the initial deployment of NGN. The countries we

evaluated initially diff ered from one another (and continue to diff er from

one another) in multiple dimensions, including: availability of alternative

last mile infrastructure, notably cable television; demography; geography

(topology); and societal and governmental attitudes toward industrial

policy.

These factors can be viewed as inputs to the formulation of overall

public policy, which drives regulatory policy. The regulatory policy that

results does not, however, uniquely determine the pace of adoption and

deployment; rather, companies base their deployment decisions on busi-

ness models that refl ect their expectations in regard to profi tability. The

regulatory framework is a factor in those expectations, but not the only

factor and not necessarily the most decisive one (Wernick, 2007). The

regulatory framework and the pace of deployment then shape the levels

of intramodal and intermodal competition going forward. Over time, the

competitive environment itself becomes an additional input to the policy

formulation process, which can be viewed as a feedback loop.

What does all of this mean? It means that at least on a worldwide scale

it is too early to expect a one- size- fi ts- all approach to the regulation of

fi ber- based NGN access. Diff erent countries are likely, for now at least, to

arrive at diff erent solutions to the range of regulatory challenges that we

have identifi ed. These diff erent approaches do not mean that the regulators

are wrong; rather, it (primarily) indicates that the circumstances of their

respective countries are diff erent from one another.22 In other words, there

is no ‘magic touch’ here, no ‘silver bullet’. There is no dominant strategy,

no solution that is clearly superior in all cases. Not yet, at least. This also

implies that a regulatory approach to the deployment of IP- based fi ber

cannot be simply and cavalierly carried over from one country to another.

In particular, US- style deregulation cannot be eff ective in a country that

does not have suffi cient alternative last mile infrastructure to build on.


The choice among the three alternatives identifi ed at the beginning of this

chapter (regulation, functional separation or deregulation) must refl ect the

market circumstances of the country or region to which it pertains.

NOTES

1. IETF, Internet Protocol: Darpa Internet Program Protocol Specifi cation, RFC 791, September 1981.

2. Indeed, technological neutrality is ostensibly a core tenet of European regulation. 3. Bitstream access seems to be much less problematic in connection with VDSL and

FTTx solutions. Also point- to- point fi ber introduces fewer problems with unbundling than PON.

4. See, for example, Van Heesvelde (2007). 5. See, for example, Amendola and Pupillo (2008) and Pileri (2008). 6. See Marcus and Elixmann (2008) for the deployment plans of Deutsche Telekom and

KPN, respectively. 7. The issue of crucial factors for the viability of an FTTx business case is also addressed

in Avisem (2007a, 2007b), ERG (2007a, 2007b) and OECD (2008a, 2008b). 8. Kirsch and von Hirschhausen (2008) consider this case. They also fi nd that NGN access

is unlikely to be replicated, that is, infrastructure competition is unlikely. 9. The issue of a potential dismantling of MDFs has been addressed in Reichl and Ruhle

(2008).10. A number of papers and reports have evaluated these costs. See Amendola and Pupillo

(2008) and J.P. Morgan (2006).11. Ofcom’s approach to risk in the assessment of the cost of capital: Ofcom (2005).12. See Marcus and Elixmann (2008) and Kirsch and von Hirschhausen (2008). KPN has

also formed a joint venture with rival telco Reggefi ber under which the pair will amal-gamate their FTTH activities in a new partnership named Reggefi ber FttH. The two companies had earlier agreed to cooperate on a FTTH project in Almere, designed to roll out fi ber- optic links to 70 000 households. The Almere project will be rolled up into the new partnership. See TeleGeography 23May 2008.

13. The number of NTT customers taking broadband services via its ADSL platform dropped from 5.3 million (as of 31 March 2007) to 4.6 million (as of 31 March 2008). The number of conventional main lines in service fell to 39.2 million from 43.3 million in this period; see TeleGeography, ‘NTT fi ber users up, ADSL subscriptions down’, 15 May 2008.

14. NTT reports that the number of people signed up to its FLET’S Hikari fi ber- optic service reached 8.78 million as of 31 March 2008, up from 3.2 million a year earlier; see TeleGeography, ‘NTT fi ber users up, ADSL subscriptions down’, 15 May 2008.

15. In October 2007, the Japanese government started to examine the interconnection charge for the fi ber access from fi scal year 2008. They decided to keep the current method, but to recalculate the fi ber cost based on high- demand estimation. As a result, fi ber charges will go down to 28.8 euros/fi ber (of NTT East).

16. Framework Directive, European Parliament and Council of the European Union (2002, p. 38).

17. DTAG has a share of 48 percent with respect to the overall (nominal) telecommunica-tions service market volume in Germany. With respect to telephony access lines DTAG has a market share of more than 80 percent (81.4 percent). DTAG has a share of slightly below 50 percent (48.6 percent) in the overall number of DSL access lines; see BNetzA (2007).

18. 1.4 million lines as of 30 June 2007 according to FCC statistics, and growing rapidly.


19. An obligation was imposed on AOL/Time Warner as a merger condition, but it proved to be ineff ective and the FCC declined to take steps to enforce it.

20. The latest FCC data are as of 30 June 2007. This time lag (nearly a year) is normal for FCC data.

21. This breakdown generally follows Kirsch and von Hirschhausen (2008).22. European regulation explicitly recognizes the principle that regulation exists to deal

with market imperfections, notably with market power, and that regulatory remedies should appropriately diff er from one country to another to the extent that market power diff ers.

REFERENCES

Amendola, G.B. and L.M. Pupillo (2008). The economics of next generation access networks and regulatory governance: towards geographic patterns of regulation. Communications and Strategies, 69, 85.

Analysys (2007a). The business case for sub- loop unbundling in the Netherlands. Study prepared for OPTA, January.

Analysys (2007b). The business case for sub- loop unbundling in Dublin. Study prepared for Comreg, December.

Avisem (2007a). Étude portant sur les modalités de déploiement d’une boucle locale fi bre optique. Study prepared for ARCEP, June.

Avisem (2007b). Étude portant sur les spécifi cations techniques des infrastruc-tures de génie civil susceptibles de supporter des réseaux d’accès FTTH. Partie 2 de l’étude: Eléments de spécifi cations des infrastructures. Study prepared for ARCEP, September.

Bundesnetzagentur (BNetza) (Germany’s Federal Network Agency) (2007). Jahresbericht 2007, http://www.bundesnetzagentur.de/enid/2.html and http://www.bundesnetzagentur.de/media/archive/13212.pdf.

ERG (2007a). ERG opinion on regulatory principles of NGA. October.ERG (2007b). Supplementary document. October.European Parliament and Council of the European Union (2002). Directive

2002/21/EC of March 7 2002 on a common regulatory framework for electronic communications networks and services (Framework Directive), (Offi cial Journal (OJ) L108, 24.04.2002, pp. 33–50). Brussels: European Parliament and Council of the European Union.

FCC (2003). In the matter of reviewing of Section 251 unbundling obligation of incumbent local exchange carriers, Triennial review order, 21 August.

Friederiszick, H.W., M. Grajek and L.- H. Roeller (2007). Analyzing the relation-ship between regulation and investment in the telecom sector. European School of Management and Technology, expert report commissioned by Deutsche Telekom, Berlin. November.

J.P. Morgan (2006). The fi ber battle. December.Katagiri, Y. (2008). Recent regulatory reform in Japanese telecommunica-

tions. Slide presentation at the International WIK Conference, Review of the European Framework for Electronic Communications. Bonn, Germany, 24–25 April.

Kirsch, F. and C. von Hirschhausen (2008). Regulation of NGN: structural sepa-ration, access regulation, or no regulation at all. Communications and Strategies, 69, 63–83.


Marcus, J.S. (2005). Is the U.S. dancing to a diff erent drummer? Communications and Strategies, 60, 4th quarter.

Marcus, J.S. and D. Elixmann (2008). Regulatory approaches to NGNs: an inter-national comparison. Communications and Strategies, 69, 19–40.

OECD (2008a). Public rights of way for fi ber deployment to the home. DSTI/ICCP/CISP(2007)5/FINAL, April.

OECD (2008b). Developments in fi ber technologies and investment, DSTI/ICCP/CISP(2007)4/FINAL, April.

Ofcom (2005). Ofcom’s approach to risk in the assessment of the cost of capital. Final statement, 18 August.

Pileri, S. (2008). Technologies and operations evolution, telecom italia analyst and investor briefi ng, 2007 Results and Strategic Guidelines. Retrieved from http://www.telecomitalia.it/TIPortale/docs/investor/Pileri_070308.pdf.

Reichl, W. and E.- O. Ruhle (2008). NGA, IP- interconnection and their impact on business models and competition. Communications and Strategies, 69, 41–62.

Taniwaki, Y. (2008). Broadband competition policy in Japan. Presentation by Ministry of Internal Aff airs and Communications (MIC), March.

Van Heesvelde, E. (2007). Approaches to VDSL regulation in Belgium. Slide presentation at the WIK Conference, The Way to Next Generation Access Networks. 21 March, Königswinter.

Wernick, C. (2007). Strategic investment decisions in regulated markets: the rela-tionship between infrastructure investments and regulation in European broad-band. Wiesbaden.

PART IV

Structural separation and regulation of the telecommunications industry

211

10. Diff usion of broadband Internet and structural separation

Arata Kamino and Hidenori Fuke

10.1 INTRODUCTION

One of the most frequently discussed topics in industrial organization

theory is vertical integration. However, the competitive eff ects of vertical

integration are two- sided. When a vertically integrated fi rm with market

power in the upstream market is also active in the downstream market,

concerns regarding competition give rise to fears that the fi rm might abuse

its market power. Abuse might take the form of obstructing competitors

with a refusal to deal or by raising rivals’ costs. To prevent this kind of

abuse by the vertically integrated fi rm, both conduct regulation and struc-

tural separation are brought into consideration. When conduct regula-

tion is deemed insuffi cient to solve the problems, structural separation is

viewed as an alternative option. However, structural separation sacrifi ces

economic effi ciency such as economies of scale and scope. In recent studies

other ineffi ciencies inherent in vertical separation such as ‘coordination’

between separated fi rms and ‘hold- up’ problems are pointed out.

In the case of the telecommunications industry, which is characterized

by dynamic technological changes, structural separation might risk rapid

transition to new technologies and preserve the market structure already

shaped by regulation. Structural separation assumes the market power of

a vertically integrated fi rm in the upstream market. However, the scope of

market power is likely to change with the introduction of new technolo-

gies. Market structure is also bound to undergo change and it becomes

diffi cult to defi ne the scope of activities which exhibit bottleneck charac-

teristics. This is especially true in the current telecommunications industry

that is experiencing a rapid transition from plain old telephone service

(POTS) to broadband. It is important, therefore, to conduct a cost–benefi t

analysis of structural separation.

Discussions on structural separation in the telecommunications sector

originated at the time of traditional POTS, an era based on the sharp dis-

tinction between long- distance and regional communications. Although


this distinction is becoming obsolete with the diff usion of the Internet,

structural separation is still on discussion tables in many of the advanced

countries except for the US. In the US, with AT&T broken up in 1984 and

divested Regional Bell Operating Companies (RBOCs) and the former

AT&T reintegrated in the mid-2000s, the issue of vertical separation is no

longer being raised. On the other hand, in Japan, where NTT was reorgan-

ized into several companies in 1999 – two regional, one long distance, one

mobile and others – all under a holding company – ownership separation

of these NTT subsidiary companies is still under consideration despite the

rapid diff usion of broadband. In the EU, meanwhile, functional separa-

tion of incumbents became a hot topic during the EC 2007 telecommuni-

cations regulatory reform debate.

This chapter tries to survey the diff erences in recent discussions on the

structural separation of incumbents in Japan, the US and Europe and

explain them from the perspective of broadband market structure in these

regions. We focus especially on Japan where ultra- fast broadband has

been developed and where the traditional approach to vertical separa-

tion needs to be reconsidered. In addition, to make our conclusions more

robust, the chapter also reviews existing market structure and discussions

on structural separation in the US and Europe. Though the situation

varies a lot among European nations, we will examine the UK, which has

introduced functional separation of British Telecommunications (BT), as

well as Germany and France as nations taking a negative position against

functional separation of their incumbent carriers.

This chapter is organized as follows. First, we will briefl y survey the

discussions on broadband market structure and structural separation.

Second, we will put forward an evaluation of the history and background

of the treatment of structural separation in Japan. Third, we evaluate

recent discussions on structural separation in the US and Europe. Finally,

based on the points above, we will explain the diff erences between struc-

tural separation policies in Japan, the US and Europe in terms of the

broadband Internet market structure and consider lessons for future

progress on the structural separation issue.

10.2 BROADBAND MARKET AND STRUCTURAL SEPARATION

One of the important issues for policy attempting to govern competition

is to constrain the abuse of market power by vertically integrated fi rms

toward the downstream market, whenever a dominant company in the

upstream market is also active in the downstream market. The measures

Diff usion of broadband Internet and structural separation 213

to secure fair competition in these circumstances are generally classifi ed

into ‘structural separation’ and ‘conduct regulation’. As incumbents in

the telecommunications industry historically dominated local facilities,

securing fair competition between incumbents and new entrants became

a big issue. Consequently, topics such as separating an incumbent’s local

businesses from its long- distance businesses (that is, structural separa-

tion) and introducing conduct regulation mainly by way of intercon-

nection policies have been widely discussed. In the telecommunications

sector, in addition to this type of narrowly defi ned structural separation,

functional separation that does not include ownership separation is

also a matter of debate. Various interpretations of vertical separation

also coexist, such as the Organisation for Economic Co- operation and

Development (OECD) distinction between ‘structural separation’ and

‘operational separation’ (OECD, 2006), or Cave’s (2006) defi nition

of eight types of separation ranging from ‘accounting separation’ to

‘ownership separation’. The European Commission (EC), in its 2007

telecommunications regulatory reform program, defi nes separation with

ownership unbundling as structural separation and separation without it

as ‘functional separation’ (EC, 2007). We will follow this EC defi nition

hereafter in this chapter. In addition, we will use ‘vertical separation’1

as including the notions of both ‘structural separation’ and ‘functional

separation’.

Vertical separation in that sense is based on the assumption that the

merits of promoting competition through regulatory measures outweigh

losses in the economies of scale and scope resulting from the subdivision

of an incumbent’s organization and/or businesses. Vertical separation,

however, raises the fundamental issue that certain forms of conduct

regulation are required as long as a discreted operating company holds

market power in the upstream market. Furthermore, vertical separation

is accompanied by the risk of preventing fl exible responses to market

changes such as the shift from traditional POTS to Internet Protocol (IP)

services. Numerous studies of theoretical industrial organization exist but

since there are various problems in applying conclusions derived from

them, policies instituted diff er signifi cantly between countries. The US

divested AT&T in 1984, and Japan reorganized NTT in 1999 by estab-

lishing a holding company overseeing a long- distance and two regional

companies, a mobile operating company, and other subsidiaries. Direct

or virtual types of structural separation as such were adopted only by

few nations in the twentieth century.2 This was to assure new entrants

of fair interconnection of long- distance networks with incumbent local

networks used for legacy fi xed voice services. In Europe, since the market

liberalization of the 1980s–1990s, there have been no cases of this kind of


vertical ‘long- distance and regional business’ separation being introduced

by regulators.

It is widely recognized that the separation of long- distance and regional

businesses is not possible and makes no sense in the recent broadband

Internet architecture. As a result, no one discusses re- enactments of

the AT&T type of divestiture. However, some European and Oceanian

countries fi nd it desirable to separate functionally the access business of

incumbents in order to remove problems with competition resulting from

the bottleneck nature of access networks sustaining broadband services.

Functional separation has already been introduced in the UK (BT), New

Zealand (TCNZ)3 and some other nations. In the US, on the other hand,

there are currently no active views on the vertical separation of incum-

bents, including functional separation of access units. In Japan, it has

been decided to resume the review of the organizational structure of NTT

in 2010. The need for resumption is based on the common recognition

among the related parties that the current divisions between local and

long- distance businesses of NTT are inappropriate in the broadband era.

Why do debates on the separation of dominant incumbents in the telecom-

munications industry diff er in Japan, the US and Europe? We suggest that

diff erences between countries derive from variations in stages and phases of

recent broadband diff usion and competition, as shown in Table 10.1.

Several offi cial broadband statistics are used in this chapter, including

the ITU Internet Report (ITU, 2006) and OECD Broadband Statistics

(OECD, 2008), providing fi gures on a global basis. To begin with, we will

briefl y examine the OECD Broadband Statistics (Table 10.2) and summa-

rize the world broadband status.

Table 10.2 reveals the characteristics of broadband diff usion in the

countries covered by this paper as follows:

Penetration rates per 100 inhabitants are almost the same among ●

Japan, the US, the UK, Germany and France (around 25 percent).

Digital subscriber line (DSL) is the most popular technology except ●

in the US and Japan. Inferring the percentages for DSL as a segment

of all broadband services, we get 95 percent for France, 93.8 percent

for Germany, 78.6 percent for the UK and 41.7 percent for Japan.

Cable ties up the majority share of all broadband services in the US ●

at 52.8 percent, while DSL occupies 40.4 percent.

Japan shows the highest (fi bre to the x) (FTTx) share with 44.3 ●

percent among all broadband access services.

(Note: These fi gures vary slightly from those in Table 10.1 mainly due to

diff erent time of records taken.)


Table 10.1 Status of broadband penetration and competition in Japan, the

US and Europe

Japan

Penetration DSL services are the fastest and cheapest in the world (ITU,

2006) and the number of subscribers to FTTx was the largest

with 14.42M at the end of 2008. The shares of FTTx, DSL

and cable modem at that time were 47.9%, 38.5% and 13.5%,

respectively (MIC, 2009a).

Competition Due to intense service- based competition, NTT’s share of retail

DSL services was 36.0% as compared to 38.2% of Softbank

Corp. at the end of 2008. NTT’s share of retail FTTx services

at the time was 73.7% (MIC, 2009b). The FTTx market is

characterized by a mixture of service- based and facilities- based

competition. KDDI, potentially the largest facilities- based

competitor to NTT, is utilizing assets acquired from electric

power companies and its own subsidiary cable TV companies.

Regulation Unbundling obligation has been imposed on both DSL and

FTTx loops of NTT.

The US

Penetration Until several years ago, the transmission speed of DSL services

was rather slow and rates were expensive compared to Japan

and Korea (ITU, 2003).1 The total number of residential high-

speed lines2 excluding mobile wireless came to 64.87M at the

end of 2007. At that point, ADSL occupied a share of 26.48M

(40.8%), cable modem 35.33M (54.5%), and fi ber 1.68M (2.6%)

(FCC, 2009). Verizon and AT&T are recently accelerating

FTTx deployment, and cable companies such as Comcast are

upgrading the transmission speed of their cable modem services.

Competition Incumbent telecommunications carriers and cable companies

are competing actively on a facilities base. ‘Triple play’

competition is intense.

Regulation FCC abolished both line sharing obligation for copper loops

and unbundling for fi bers excluding voice 64 Kbps path in

2005.

Europe (EU)

Penetration Penetration rates vary considerably among EU member states,

from the Slovak Republic with the lowest rate at 8.9% to


Table 10.1 (continued)

Denmark with the highest at 36.7%. However, many of the

major member states show a penetration rate almost the same

as in Japan (23.0%) and in the US (25.0%). It is 27.6% in the

UK, 26.4% in France and 26.2% in Germany (OECD, 2008).

Except for a few nations where cable modems are dominant

and Sweden where FTTx is widely deployed, the dominant

broadband technology is DSL. However, commercial FTTx

services are starting to be launched in Germany and France.

Competition Competition in many of the member states is limited to

intramodal service- based DSL. To promote service- based

competition, local loop unbundle (LLU) and bit stream access

(BSA: a kind of wholesale product) are being implemented.

The shares of incumbents in retail markets and the usage rates

of LLU and BSA vary signifi cantly among member states.

Regulation The EU has shifted from the unifi ed application of The LLU

Regulation (EC, 2000) to the introduction of LLU and BSA

obligations, a process which was based on market analysis in

member states under the 2003 regulatory framework. Most

states impose LLU and BSA obligations on incumbents. As

for FTTx, many countries are conducting a regulatory review

to decide how to treat FTTx. Remedies such as open access to

duct as the fi rst step, with dark fi ber provision as the second

step and sharing of intra- building wire are under discussion.

Notes:1 Prices per 100 Kbps were $0.09 (Japan), $0.25 (Korea) and $3.53 (the US) at the time.2 Over 200 Kbps in at least one direction.

Table 10.2 Broadband penetration rates per 100 inhabitants by

technology in major countries as of June 2008

Penetration DSL (%) Cable (%) FTTx/LAN

(%)

Others (%) Total (%)

Japan 9.6 3.1 10.2 0.0 23.0

The US 10.1 13.2 0.9 0.8 25.0

The UK 21.7 5.9 0.0 0.1 27.6

Germany 24.6 1.6 0.0 0.1 26.2

France 25.1 1.3 0.0 0.0 26.4

Source: OECD (2008).


10.3 PROGRESS ON BROADBAND COMPETITION AND STRUCTURAL SEPARATION IN JAPAN

10.3.1 Telecommunications Liberalization and Competition Policy

In Japan, the liberalization of the telecommunications market and the pri-

vatization of NTT (Nippon Telegraph and Telephone Public Corporation)

were implemented simultaneously in 1985. However, as Fuke (2007)

points out, there was no suffi cient debate on whether conduct regula-

tion should be imposed on NTT or whether structural separation should

be chosen in order to secure fair competition. Though many European

countries arranged interconnection rules and rate rebalancing at the time

of market liberalization around 1998, interconnection rules were not well

provided for in Japan until 1997 when the Ministry of Internal Aff airs

and Communications (MIC) implemented new rules by revising the

Telecommunications Business Law. One of the reasons for the diff erence

between Japan and the EU is the fact that market liberalization occurred

13 years earlier in Japan than in the EU, with very few international

benchmark rules applicable at that time. Since the initial arrangement of

conduct regulation was insuffi cient in Japan as already mentioned, views

have continued to be canvassed on the structural separation of NTT into

long- distance business and regional business as a tool to promote fair

competition. Thus the structural separation of NTT has been under con-

sideration in parallel with the arrangement of conduct regulations in the

1990s. The persistence of these discussions has been one of the particular

characteristics of the Japanese telecommunications market to this day.

As a result, Japan introduced an indirect but practical structural sepa-

ration of NTT. NTT was reorganized under a holding company structure

in 1999, while adopting the same level of conduct regulations as the EU.

However, it became clear that such a ‘regional versus long- distance’ type

of structural separation could not keep up with massive telecommunica-

tions changes from POTS to the Internet and to mobile communications.

In order to resolve this problem, MIC (2006) issued a ministerial report in

2006 that proposed the removal of line of business control of NTT regional

companies. This was to be achieved by instituting additional fair competi-

tion rules including the ownership separation of NTT group companies

by way of disbanding the NTT Holding Company in 2010. However,

due to a lack of a concrete vision for NTT after reviewing its organiza-

tion and given the need for further conduct regulations, the Cabinet and

ruling party (that is, the Liberal Democratic Party) at the time decided

to suspend the MIC (2006) panel proposal and agreed to postpone the

discussions on NTT reorganization until 2010. The agreement said that:


‘as for the organizational structure of NTT, we shall resume the review in

2010, taking into account the status of broadband penetration and NTT

Group’s Medium- Term Management Strategy’.4 Conclusions are sup-

posed to be reached as soon as possible following the future review.

10.3.2 Service- Based Competition and Facilities- Based Competition

Since the end of the 1990s, Japan has been implementing the most rigid

open network policy among major developed countries to promote

service- based competition. For example, unbundling obligation has con-

sistently been imposed on both copper and fi ber loops until now. There is

no bit stream access (BSA) type of broadband wholesale scheme because

unbundled- based competition relying on very low local loop unbundling

(LLU) charges (especially for line sharing) has fl ourished from the begin-

ning and there has been no room for Internet service providers (ISPs) to

adopt BSA- based competition. How does one explain that the US and

European countries, which implemented more or less the same kind of

copper LLU as Japan, did not experience the same degree of DSL compe-

tition as Japan? There might be no single answer to this question (Ikeda,

2005). As the LLU charges were set very low, new entries by service- based

carriers were encouraged. Furthermore, there have been competitors who,

driven by a kind of ‘animal spirit’ (Akerlof and Shiller, 2009), continued

to provide very cheap alternative DSL services despite long- standing fi scal

losses during early 2000s.

Table 10.3 shows that line sharing charges were set very low during

that period in Japan. Nevertheless, Softbank Corp., the largest DSL

competitor, incurred a huge loss (Table 10.4) because the company set a

disruptively cheap retail rate for DSL and, in the short term, faced massive

marketing costs in capturing the majority market share (Kamino, 2007).

As the International Telecommunication Union (ITU) (2003) observed,

a strategy like this contributed to the realization of ‘the world fastest and

cheapest DSL’.

Table 10.5 summarizes characteristics of DSL competition in Japan.

Though these characteristics have both merits and demerits, it is probable

that demerits exceed merits in some cases. Further analysis will be required

to make a full assessment of the balance between them to judge the result

of DSL competition in Japan.

Eventually, Japanese competitive broadband providers would not have

been able to sustain losses in FTTx businesses in the same manner in

which DSL services had initially recorded substantial losses (Table 10.4).

Consequently, unwilling to fall into the same money- losing trap, they

requested regulation enabling very low wholesale charges for fi ber loops.


It is precisely this kind of regulatory interference of setting very low whole-

sale prices that is likely to lead to a regulatory preference for a specifi c

technology (that is, FTTx), which, in turn, might impair the development

of other broadband technologies.

10.3.3 Status of Broadband Penetration and Competition

The Japanese broadband market entered the FTTx era prior to other coun-

tries. Given the rapid expansion of the market, the number of FTTx sub-

scribers exceeded that of DSL in the fi rst fi scal quarter of 2008 (April–June)

(Figure 10.1).5 As noted above in terms of intramodal competition within

DSL, the total NTT market share of FTTx (73.7 percent) is considerably

higher compared to that of DSL, with some exceptional competitors such

as K- Opti.com, a subsidiary communications company of Kansai Electric

Company in Osaka, holding the majority FTTx share in several prefec-

tures6 (MIC, 2009b). In a situation where even NTT itself is often asked

Table 10.3 Comparison of LLU charges (October, 2004)

Full unbundling Shared access (line sharing)

Japan NTT East: 1,256 yen

NTT West: 1,318 yen

NTT East: 158 yen

NTT West: 165 yen

The UK 8.76 pound

(1,752 yen)

2.26 pound

(452 yen)

Germany 11.80 euro

(1,652 yen)

2.43 euro

(340 yen)

France 10.50 euro

(1,470 yen)

2.90 euro

(406 yen)

Note: Calculated on the basis of the exchange rate at the time (1 pound = 200 yen, 1 euro = 140 yen).

Source: NTT tariff , BT Price List, Ofcom (2004) and RegTP (2004).

Table 10.4 Financial result of Softbank Corp.’s broadband infrastructure

(million yen)

Fiscal year 2003 2004 2005 2006 2007

Operating revenue 40 007 128 906 205 306 268 451 258 824

Operating profi t −96 204 −87 597 −53 747 20 672 26 809

Source: Softbank Corp.’s fi nancial data


the reason why it is making huge investments in currently less profi table

FTTx deployment, it is certainly diffi cult for competitors to shift wholly

into defi cit- making service- based FTTx competition, just as it had been

the case with DSL (Table 10.4). While facilities- based competition between

NTT and electric companies that have large access fi ber assets was once

thought to be feasible, electric companies, with the exception of some rare

cases such as K- Opti.com, are reluctant to keep on promoting FTTx busi-

ness as they had been doing before. In spite of this, with KDDI’s purchase

of a subsidiary communications company and other fi ber access businesses

from Tokyo Electric Company (TEPCO), signs are emerging that facilities-

Table 10.5 Characteristics of DSL competition in Japan

Characteristics Merits Demerits

(1) Very low LLU

charges

Intense LLU- based

competition developed.

Neither intramodal

facilities-based

competition nor

intermodal competition

developed.

(2) Destructively

cheap DSL retail

prices

Only large companies

with fi nancial

abundance could

bring sustainable

competition.

FTTx retail price was

set very low because

cheap DSL prices had

become benchmark for

broadband prices.

Only competitors with large

fi nancial base that could

bear defi cits over several

years were able to enter

the market.

Margin squeeze regulation

applied to FTTx

wholesale charges

restricted the development

of facilities- based

competition (Fuke, 2007).

(3) Fierce DSL

competition

NTT shifted its

broadband strategy

from DSL, in which

it has less market

share and profi t, to

FTTx. This, in turn,

accelerated FTTx

deployment and

penetration.

While DSL competitors

faced a net decrease

in the number of DSL

subscribers due to

customer shift to FTTx,

they started to produce

profi ts in DSL businesses.

(DSL customer numbers

hit the peak in March

2006 and kept on

declining after that.)


based competition between NTT and other telecommunications providers

such as KDDI is occurring in some densely populated areas.

As for cable broadband, which is widely operated in some countries

including the US, the household penetration rate for cable TV in Japan

soared to 33 percent around 2003 and then kept rising to 42 percent by the

end of 2007. Nonetheless, cable broadband market share within the whole

range of broadband technologies is slackening, at around 15 percent

between 2005 and 2007. In summary, the characteristics of the Japanese

cable industry can be listed as follows (MIC, 2008):

Deregulation for multiple system operator (MSO) restriction was ●

delayed until 1993, and MSOs less large than expected have appeared

since then.7 Though Jupiter Telecommunications Co. (J:COM),

the largest MSO, was established in 1995, its subscribers came to

only about 2.18 million by the end of February 2008. This number

appears considerably small compared with 24.1 million at the end of

2007 for Comcast, the largest MSO in the US.

Although the household penetration rate for cable TV itself has been ●

increasing, overall cable broadband penetration per 100 inhabitants

16000

14000

12000

10000

8000

6000

4000

2000

2000

0

3 9

2001

3 9

2002

3 9

2003

3 9

2004

3 9

2005

3 9

2006

3 9

2007

3 9

2008

3 9

Th

ou

san

ds

Cable

DSL

FTTH

FWA

Source: MIC (2009a).

Figure 10.1 Growth of broadband services in Japan


has been decreasing slightly, from 15.4 percent to 13.5 percent over

the three- year period 2005–07.

As cable broadband does not play a signifi cant role in triple play ●

competition, the number of subscribers to IP multicast- type broad-

band broadcasting (that is, Internet Protocol television – IPTV)

remained sluggish with just 240 000 subscribers at the end of 2007.

As we have seen, it was unfortunate for the cable industry that service-

based DSL competition was so intense that there was little room for cable

broadband to expand. Development of the Japanese broadband market

is summarized in Figure 10.2. Both MIC’s regulatory policy to enforce

low LLU charges and the aggressive market entry of Softbank Corp. con-

tributed to the rapid deployment of cheap and fast DSL services.8 NTT,

having been left behind by Softbank Corp. in the DSL market, has tried to

catch up by focusing on FTTx services. However, it was forced to set retail

prices of FTTx markedly low to enable it to compete against Softbank

Corp.’s cheap DSL as well as against the FTTx services of newly entered

electric power companies. By setting retail prices low, NTT was then

obliged to set wholesale charges of FTTx below any profi table level due to

margin squeeze regulation (Fuke, 2007).

Even with these low wholesale FTTx charges, service- based competitors

have not been able to expand their retail FTTx. It has been diffi cult for

them to invest in it due to the fi nancial burden of their DSL businesses,

where they have not fully recovered accumulated losses. Thus NTT’s share

in the FTTx market has continued to rise from 57.5 percent to 73.7 percent

Agressive Marketing by Softbank Corp.Low LLU & Co-location Charges

Very Cheap and Fast DSL Services

Cheap Retail FTTx Services by Electric Power Companies

Cheap Retail FTTx Services by NTT

Low Wholesale FTTx Charges by NTT

• Weak service-based competition in FTTx market

• Limited facilities-based competition in FTTx market

• Weak inter and intramodal competition in broadband market

MIC’s Margin

Squeeze Regulation

MIC’s Regulation

Figure 10.2 Infl uence of severe DSL competition on FTTx


through the years 2005–08 (MIC, 2009b), with some competitors again

arguing for the structural separation of NTT.

10.3.4 Current Progress on Vertical Separation

As shown below, progress on structural separation of NTT in Japan has

aspects quite diff erent from those in other countries:

Current discussions in Japan followed the debates on ownership ●

separation of long- distance and regional NTT businesses, originally

started in the 1980s.

Fundamental issues have not yet been resolved even following 1999 ●

when a compromise was reached to reorganize NTT under a holding

company structure.

Reviews undertaken during 2005–06 to resolve these issues were ●

suspended as a result of political compromise.

As was pointed out by the MIC Minister’s panel report (MIC, 2006),

NTT has been constrained by line- of- business control regulations that do

not accommodate current technological and market environments. The

problem arose because current laws and regulations designed for the reor-

ganization of NTT in 1999 did not take into account rapid technological

innovation. It appears unreasonable to maintain the current regulations,

which virtually separate long- distance and regional NTT businesses (MIC,

2006).

Recent progress in the broadband market, especially that of DSL and

FTTx services that were reviewed above, might support the argument for

reorganization (that is, structural separation) of NTT. However, it must

be noted that this market situation is the result of MIC’s regulation in

favor of service- based competition. If we want to encourage the develop-

ment of facilities- based competition in the broadband market, structural

separation is not a unique option. We should rather revise the regulatory

policy that is much biased toward service- based competition.

10.4 CURRENT DISCUSSIONS ON VERTICAL SEPARATION IN THE US AND EUROPE

10.4.1 The Case of the US

There is little debate on vertical separation in the US as of early 2009.

Large mergers and acquisitions (M&A) cases between RBOCs took place


around 2005–06 and since then AT&T and Verizon have come to dominate

the telecommunications market. The main reviewers of those merger plans

were the Department of Justice (DoJ) and the Federal Communications

Commission (FCC). The FCC conducted an analysis balancing ‘potential

harms and benefi ts to public interests’ in its merger approval procedures.

The FCC then judged that merger benefi ts outweighed harms if certain

attached conditions were implemented. As will be seen below, FCC’s ana-

lytical framework is built upon the balancing theory of the post- Chicago

School approach (Yoo, 2002).

On the one hand, during the proposed SBC–AT&T and Verizon–MCI

merger reviews, the specifi c markets (for example, ‘enterprise/mass market

retail services’ and ‘the Internet backbone market’; FCC, 2005) that might

have been harmfully aff ected by anticompetitive eff ect were observed. On

the other hand, four potential benefi ts were considered by the FCC, as

shown below:

enhancements of national security and government services; ●

effi ciencies related to vertical integration; ●

economies of scope and scale; ●

cost synergies. ●

Considering vertical separation, it should be noted that the second and

third factors were regarded as balancing factors promoting public inter-

ests. As Riordan (2008) pointed out, the FCC drew this conclusion based

on a more realistic ‘give and take negotiation theory’ of the post- Chicago

School thinking that the FCC had been adopting since the News Corp.

and DirecTV merger case.

No discussions on vertical separation are currently taking place. If we

try to fi nd any sort of debate on vertical separation, it will involve the

type of questions associated with merger cases such as: Would the costs

of further strengthening vertically integrated incumbent carriers outweigh

its benefi ts? This approach lies within the scope of post- Chicago School

theory. However, even under a Republican- led FCC, as shown in the

AT&T–BellSouth merger review, commissioners from both Democrat and

Republican circles voiced considerably diff erent opinions on balancing

merits and demerits of the merger. As a result, additional voluntary com-

mitments such as augmented net neutrality promise are required for the

merger plan applicants (AT&T–BellSouth).

Generally speaking, there is a widespread tendency in the US to conduct

analysis by balancing merits and demerits of vertical integration while

taking into account particular situations. This trend is clearly found in

recent reviews of the US M&A cases (Riordan, 2008). We should point out


that the non- existence of arguments for vertical separation in US telecom-

munications businesses in recent years is a refl ection of the development

of facilities- based competition in the broadband Internet market between

telecommunications companies and cable operators. This unique market

structure is also the result of the regulatory environment, including ‘fi n-

syn’ (fi nancial interest and syndication) rules, which applies to the broad-

cast and media industry.

10.4.2 The Case of the European Union (EU)

Pros and cons of functional separation have been widely discussed in

Europe (Waverman, 2006). It is true that LLU- based competition had

not progressed much in the UK’s DSL market at the time Ofcom recom-

mended functional separation of BT (Ofcom, 2005) and accepted BT

Undertakings (BT, 2005) in its Telecommunications Strategic Review

(TSR). It is equally true that taking up of LLU has been accelerated since

BT established its Openreach access division. However, Germany and

France, which had not introduced any kind of functional separation, had

already experienced explosive use of LLU when Ofcom initiated TSR in

April 2004 (Table 10.6). Following the reorganization of NTT in 1999, the

Japanese DSL market started to show spectacular growth from 2000 when

LLU and collocation were formally mandated.9 These facts lend little

support to the assumption that functional separation itself promotes LLU

usage. Therefore, it would be reasonable to think that necessary and suffi -

cient conditions for the promotion of broadband competition are brought

about not by functional separation but by equivalence of input (EOI) and

fair rate settings for wholesale inputs.

Since functional separation was imposed on BT, the number of coun-

tries considering the introduction of the same kind of functional sepa-

ration has been increasing. The EC is the most active supporter of this

Table 10.6 Take- up of LLU in the UK, Germany and France (end of

June, 2004)

Country Full LLU Shared

access (line

sharing)

Total LLU

DSL use Voice use Total

UK 7 580 0 7 580 6 270 13 850

Germany 650 000 978 699 1 628 699 0 1 628 699

France 13 066 0 13 066 717 654 730 720

Source: ECTA (2004).


trend and took up functional separation as a ‘remedy of last resort’ in the

proposals for its revised directives in the 2007 telecommunications regula-

tory reform (EC, 2007). However, key related parties opposed structural

separation and the EC itself revealed an opinion recommending functional

separation rather than structural separation in an explanatory note of

the proposed EU directives, concluding that: ‘given this experience and

the high level and non- revocable intervention involved, very signifi cant

benefi ts of mandated structural separation in terms of gains from achiev-

ing equality would have to be demonstrated for it to be a suitable remedy

in the telecommunications sector’ (EC, 2007). Even among EC commis-

sioners, some voices opposed to functional separation pushed by Telecom

Commissioner Ms Reding could be heard before the proposals were sub-

mitted to the European Parliament and the Council. This led Competition

Commissioner Ms Kroes and Enterprise and Industry Commissioner Mr

Verheugen to criticize strongly Ms Reding’s idea according to articles

from the Financial Times excerpted below:

Neelie Kroes, EU Competition Chief, and Günter Verheugen, the Industry Commissioner, have attacked a proposed overhaul by fellow Commissioner Viviane Reding of laws governing the bloc’s €289bn- a- year ($407bn, £201bn) electronic communications sector. They warn that the proposed review could create more bureaucracy and harm investment . . . Ms. Kroes’ offi cials argue that this move – known as ‘functional separation’ – risks hitting investment in the sector, especially in new, ultra- fast broadband networks. An internal docu-ment written by her department, and seen by the FT, said the measure ‘is not only superfl uous but also damaging. Functional separation does not prevent discrimination of alternative operators’.10

The European Regulators Group (ERG) announced acceptance of the

functional separation proposal in October 2007 with the provision that the

fi nal decision on the introduction of functional separation is solely that of

each member state (ERG, 2007). Even so, some NRAs (national regulatory

authorities) such as ARCEP (France) or CMT (Spain), who participated in

the ERG announcement, expressed their intention to refuse the adoption

of functional separation in their own countries. Again, some government

offi cials such as two vice- ministers in the German Ministry of Economics

and Technologies disallowed the functional separation remedy itself. On

the other hand, several countries such as Sweden and Italy are progressing

towards the introduction of functional separation. Tensions are arising

there between the EC, NRAs and incumbents because the targeted incum-

bents (Telia Sonera and Telecom Italia) voluntarily established new access

business units to avoid regulatory mandated functional separation.

Though the EU scheduled the completion of deliberations on the 2007


telecommunications regulatory reform by the end of 2009 at the latest,11 it

was expected that many EU countries would suspend further considera-

tion on functional separation until then. They will also draw conclusions

on the eff ectiveness of functional separation from the case analysis of BT,

given that it has already been more than three years since BT established

Openreach in January 2006.

Incumbent carriers still hold strong market power in many of the

European countries, a fact which refl ects the rather late introduction

of full competition in telecommunications in 1998. Broadband services

started to evolve before competition had fully unfolded in the POTS

market. Facilities- based competition is not likely to develop in these coun-

tries in the near future. In evaluating discussions on vertical separation, it

is necessary to take into account this stage of market development.

Here it is useful to study Cave’s suggestion, from the perspective of

market reality, that:

it is important that this remedy [of vertical separation] be applied proportion-ately. This requires that the detriments resulting from non- price discrimination exceed the costs of imposing an operational separation remedy, where those costs are not only those of changing the incumbents’ business processes, but also [those] of any chilling eff ect on investment in new assets, by both the incumbent and competitors. (Cave, 2006)

At the same time, we should remember experiences made in Japan and

France, which indicate that eff ective conduct regulation for mandating

LLU and collocation, rather than functional separation, can lead to rapid

growth in the DSL and total broadband market.

10.5 CONCLUDING ANALYSIS AND FURTHER ISSUES

10.5.1 Concluding Analysis

Given the above analysis of the broadband market and regulatory policy

aff ecting it, what implications does vertical separation policy have for the

three regions under consideration, Japan, the US and Europe? Our assess-

ment is that current diff erences in the discussions on vertical separation

are a refl ection of diff erences in the market structure of broadband and

that the market structure itself is formed by regulatory policy. In assessing

the current status of deliberations on vertical separation, it is therefore

necessary to understand the interaction between market structure and

regulatory policy.


In the US, political and regulatory confusion deterred the deployment

of DSL from the start and non- regulated cable modem services were able

to spread ahead of DSL. To make up for this ‘hold- up’, the integrated

incumbent operators focused on FTTx deployment alongside DSL, and

such activities led to facilities- based competition in the broadband market.

As a result, it became unnecessary to implement vertical separation of

incumbents. In other words, we should like to make the point that the

failure of regulatory policy resulted in an unexpected outcome which ren-

dered the need to impose regulation unnecessary.

Though NTT was reorganized in 1999 to promote competition in tra-

ditional POTS, contradictions have been coming up from that time with

the rapid growth of broadband Internet. NTT regional companies’ launch

of commercial NGN services in March 2008 may serve as an example.

The way NTT provides NGN reveals the contradiction inherent in the

regulation limiting NTT regional companies’ business areas within intra-

prefecture communications. As the Internet has developed as a global

seamless network, there ought to be no boundary between long- distance

and regional communications. When NTT regional companies tried to

off er NGN, they were required to off er the intra- prefecture parts of the

service themselves while leasing inter- prefecture parts from other carriers.

That is to say, regulation forced an artifi cial division on the Internet. How

should we deal with ‘vertical separation problems’ based on these market

trends? It is important to recognize that the current market status is also

a product of regulation. While it is true that LLU obligation mandated

by MIC contributed to the rapid growth of the DSL market, it is service-

based competition that relied on local loops of NTT regional companies.

Though it seemed that facilities- based competition would grow in the

fi ber to the home (FTTH) market initially, MIC’s policy of setting dark

fi ber charges at a low level restrains the development of facilities based-

competition not only in FTTH itself but also in cable broadband and wire-

less services such as WiMAX (Worldwide Interoperability for Microwave

Access) (Fuke, 2007).

In this situation, competition in the Japanese broadband market might

remain service- based and no signifi cant facilities- based competition may

materialize. In general, ownership separation of an incumbent carrier or

functional separation of its bottleneck unit would not lead to facilities-

based competition even if it promoted service- based competition. Clearly,

facilities- based competition has an advantage over service- based competi-

tion in the sense that it stimulates incentives for technological innovation

while ownership separation on its own will not lead to facilities- based

competition.

Finally, how should Europe treat vertical separation policy given the


lessons of Japanese and the US experiences? Our chapter takes the view

that the fi nal goal in this debate is to materialize facilities- based competi-

tion. It is eff ective to impose LLU and lower rate setting for LLU in order

to promote competition in the short term. However, as shown above, verti-

cal separation as such is not eff ective. It can be anticipated that incumbent

operators would show reluctance to accept a strict unbundling regulation.

In this case, it may be more eff ective to show intentions of introducing ver-

tical separation as a kind of ‘stick’, in turn inducing incumbents to accept

regulation.

Although this kind of remedy would lead to service- based competition,

it might hamper facilities- based competition as shown in the case of Japan.

Hence, it should be regarded as a temporary remedy until an adequate

degree of competition develops in the broadband market. In any case,

regulators should hesitate before enforcing vertical separation without

taking into account its inherent demerits.

10.5.2 Further Issues

The introduction of FTTx and NGN built on FTTx raises additional issues

in the discussion of vertical separation in telecommunications industry.

The fi rst issue is related to the ‘hold- up’ and ‘coordination’ problem. In

the case of POTS, technology is mature and it is easy to defi ne bottleneck

parts of the network. However, FTTx is being newly built and technology

is going through innovation. For example, NTT is utilizing various tech-

nologies for FTTx including fi ber to the building (connecting homes with

copper intra- building wires) and shared use of a fi ber core from a local

exchange to a nearby cabinet shared by several users. In these cases, it is

diffi cult to tell which part of a fi ber might constitute a bottleneck. Further,

the build- out of fi bers and the adoption of a particular technology depend

on the forecast of user demand in terms of service quality and quantity.

However, the fi rm in charge of maintaining access facilities cannot obtain

the kind of user information collected by the fi rm in charge of providing

retail services relied on wholesale products. Even if the upstream fi rm

invests in facilities based on a particular technology, there is no guarantee

that the fi rms in the downstream market will lease the wholesale facilities.

In these circumstances, the upstream fi rm will become too conservative

in new investment and is likely to hamper the smooth deployment of

new services. When we discuss vertical separation in telecommunications

industry, we should therefore keep in mind this kind of ‘hold- up’ and

‘coordination’ problem.

The second issue is related to content and application providers.

Current discussions on vertical separation are mainly focusing on fair


competition between vertically integrated telecommunications carriers

and non- integrated telecommunications carriers. Since non- integrated

competitors usually lack access assets, current discussions on vertical

separation are heating up on the topic of functional separation of access

bottlenecks. In the age of core NGNs and NGA (next generation networks

and next generation access), however, competition between telecommuni-

cations carriers and content providers many of whom do not own network

facilities (that is, ultimately non- integrated providers) is expanding. In

cases where free broadband content is subsidized with advertising revenue

from content providers and is competing with content by new entrants

from the telecommunications market, it is not clear at present that owning

the network (that is, holding vertically integrated organization) is always

accompanied by competitive advantages. In particular, it can be assumed

from net neutrality discussions that, if incumbents are unable to change

their current rate structure for broadband access provision, upgrading,

maintaining access and combined backhaul, then operating a network

may become burdensome for them. Therefore, future discussions on

vertical separation should consider the perspective of competition that is

emerging between traditional carriers and content providers on the basis

of new business models.

NOTES

1. Theories on vertical separation in telecommunications industry are analyzed in detail by Sasaki (2007).

2. Though NTT keeps ownership relationships between long- distance, regional, mobile and other businesses under a holding company, such an organizational structure is the tentative result of deliberations on the structural separation of NTT that have continued since the privatization of the company. Each operating company is regulated separately to conduct businesses stipulated by NTT Law and by a guideline issued at the time of reorganization in 1999. Therefore, we should regard current NTT group companies as virtually divested companies.

3. TeliaSonera and Telecom Italia voluntarily established functionally separated access divisions in early 2008.

4. NTT published its Medium- Term Management Strategy in November 2004 and prom-ised to shift 30 million customers to optical fi ber access and next generation network services by 2010 (NTT, 2004). The promise has been modifi ed several times after then.

5. Status of broadband diff usion and competition in Japan is described and analyzed in detail by Fuke (2008).

6. Japan’s local administration is divided into 47 prefectures. 7. Since it was permitted for one cable operator to extend their businesses in several

separated regions and to accept investment from foreign companies, MSOs started to emerge.

8. See Kamino (2007) for prehistory of Softbank Corp.’s entry into the broadband market.

9. See Fuke (2007) and Fuke (2008) for details. Metrical analysis by Akematsu and Tsuji


extracts regulatory policy, competition and technological development as factors for DSL progress in Japan (Akematsu and Tsuji, 2006).

10. It is cited from an article from the Financial Times titled ‘Brussels divided over telecoms plan’ (25 September 2007).

11. The reform was formally appraised in late November 2009 and entered into force on 19 December 2009.

REFERENCES

Akematsu, Y. and M. Tsuji (2006). Deregulation or market competition, which has larger eff ect on Japanese ADSL development; panel data and AHP analyses. Paper presented at ITS 17th Biannual Conference.

Akerlof, G. and Shiller, R. (2009). Animal Spirits: How Human Psychology Drives the Economy, and Why It Matters for Global Capitalism. Princeton, NJ: Princeton University Press.

BT (2005). Undertakings given to Ofcom by BT pursuant to the Enterprise Act 2002. 22 September.

Cave, M. (2006). Six degrees of separation: operational separation as a remedy in European telecommunications regulation. Communications and Strategies, 64, 89–103.

EC (2000). The European Parliament and the Council of 18 December 2000. Regulation (EC) No 2887/2000 on unbundled access to the local loop.

EC (2007). Commission staff working document. Impact assessment. Accompanying document to the Commission proposal for a Directive of the European Parliament and the Council, {COM (2007) 697, COM (2007)698, COM (2007)699, SEC (2007)1473}, 13 November.

ECTA (2004). ECTA broadband scorecard end of June 2004. 28 September.ERG (2007). ERG opinion on functional separation. ERG(07)44.FCC (2005). In the Matter of SBC Communications Inc. and AT&T Corp.

Applications for Approval of Transfer of Control, Memorandum Opinion and order, FCC 05- 183, 17 November.

FCC (2009). High- speed service for Internet access: status as of December 31, 2007.

Fuke, H. (2007). Info- Communications Policy in the Broadband Era. Tokyo: NTT Publishing. (In Japanese.)

Fuke, H. (2008). Structural changes and regulatory challenges in Japanese telecom-munications industry. In Y. Dwivedi, A. Papazafeiropoulou and J. Choodrie (eds), Handbook of Research on Global Diff usion of Broadband Data Transmission Volume 1 (pp. 90–107). Hershey, PA: Information Science Reference.

Ikeda, N. (2005). The Architecture of Information Technologies and Organizations. Tokyo: NTT Publishing. (In Japanese.)

ITU (2003). ITU internet report 2003: birth of broadband.ITU (2006). ITU internet report 2006: digital life.Kamino, A. (2007). ‘Realization of complementarities through full scale integra-

tion – M&A in telecommunications sector’. In H. Myiajima (ed.), M&A in Japan (pp. 259–82). Tokyo: Toyo Keizai. (In Japanese.)

MIC (2006). Report of Panel on Frameworks of Communications and Broadcasting. (In Japanese.)


MIC (2008). Status of cable TV as of end December 2007, February 2008. (In Japanese.)

MIC (2009a). Number of Broadband Service Contracts, Etc. (as of the end of December 2008).

MIC (2009b). Disclosure of Quarterly Data concerning Competition Review in the Telecommunications Business Field, Third quarter of FY2008 (as of the end of December 2008).

NTT (2004). NTT Group’s medium- term management strategy. Press release, 10 November. Retrieved from http://www.ntt.co.jp/news/news04e/0411/041110d.html.

OECD (2006). Report to the council on experiences on the implementation of the recommendation concerning structural separation in regulated industries, (C(2006)(65).

OECD (2008). OECD Broadband Statistics, OECD Broadband Portal, June.Ofcom (2004). Review of the wholesale local access market Explanatory statement

and notifi cation, 26 August.Ofcom (2005). Final statements on the Strategic Review of Telecommunication,

and undertakings in lieu of a reference under the Enterprise Act 2002, 22 September.

RegTP (2004). RegTP Newsletter. One- off local loop charges. 25 June.Riordan, M. (2008). Competitive Eff ects of Vertical Integration. In P. Buccirossi

(ed.), Handbook of Antitrust Economics (pp. 145–82). Cambridge, MA: MIT Press.

Sasaki, T. (2007). Perspective on structural separations or functional separations. Working paper No. 07- 02, August. (In Japanese.)

Waverman, L. (2006). The challenges of a digital world and the need for a new regulatory paradigm at Ofcom. In Ofcom (edited by E. Richards, R. Foster and T. Kiedrowski), Communications – The Next Decade, A Collection Of Papers Prepared for the UK Offi ce of Communications, London: Ofcom, pp. 158–75.

Yoo, C. (2002). Vertical integration and media regulation. Yale Journal on Regulation, 19, 171–300.

233

11. Implementing functional separation in fi xed telecommunications markets: the UK experience

Peter Curwen and Jason Whalley

11.1 INTRODUCTION

Since 2004 or so there has been considerable interest in the implementa-

tion of functional separation within fi xed telecommunications markets. In

these markets, which are largely but not exclusively to be found within the

European Union, functional separation is seen as a way to resolve the ten-

sions that exist between incumbent operators and those other service provid-

ers that require access to incumbents’ networks to deliver their own services.

At the forefront of the implementation of functional separation is the

UK. In late 2005, Ofcom and British Telecom (BT) agreed on a series

of undertakings that culminated in the creation of a new company,

Openreach, to run BT’s local access network. Accompanying the estab-

lishment of Openreach was the imposition of key performance indicators

and penalties for non- achievement. As Openreach is clearly an important

landmark in the development of the UK telecommunications market, this

chapter will focus on the implementation of the undertakings to date.

With this in mind, the remainder of this chapter is divided into fi ve sec-

tions. A brief overview of the diff erent types of separation possible within

the telecommunications industry is provided in section 11.2. Background

information regarding the adoption of functional separation is detailed

in section 11.3, while section 11.4 focuses on the implementation of

the undertakings. The adoption of functional separation in the UK is

appraised in section 11.5, and conclusions drawn in section 11.6.

11.2 LITERATURE

Any assessment of functional separation within the UK raises two ques-

tions: What is functional separation, and what has been the UK experience?


This section will address the former of these questions, while the remain-

der of the chapter will address the second question. A useful starting point

when answering the fi rst question is Xavier and Ypsilanti (2004). Through

focusing on the separation of competitive from non- competitive services,

the authors (ibid., p. 76) identify a range of separation measures that have

been implemented within the telecommunications industry, namely:

accounting, functional and corporate separation; ●

separation into regional operators; ●

separation of local from long- distance services; ●

separation of local and mobile services; ●

separation of local and broadband/advanced services; ●

separation of an incumbent into smaller, vertically integrated ●

carriers.

With respect to structural separation, four diff erent approaches are

identifi ed (ibid., pp. 77–81). In the fi rst of these, LoopCo, the incumbent

divests its access business to form a new company while in the second,

NetCo, an arm’s- length relationship is established between the incum-

bent’s access and non- access networks. The establishment of an alterna-

tive distribution company is the third approach suggested, and entails

the collective ownership of the non- competitive assets by those operators

present in the competitive parts of the market. The fi nal approach sug-

gested is that of voluntary suspension.1

Although all four approaches have been variously discussed, it is not

altogether surprising that attention has focused on the LoopCo and NetCo

proposals. Cave (2002), Dounoukos and Henderson (2003) and Xavier

and Ypsilanti (2004) identify a range of factors that need to be taken into

consideration regarding the viability of the LoopCo option. One such

factor is the network scope of LoopCo, while another is whether coordi-

nation problems between LoopCo and other operators would result as a

consequence of technological advances. With respect to the restructuring

of BT announced in late 2000, Sandbach (2001) stated that the proposed

NetCo inadequately addressed the local loop issues as it would cover both

the copper and switched network. That is, the possibility remained for

NetCo to favor its own services or those provided elsewhere in BT.

Cave (2002, p. 30) argues that with the NetCo option there is the pos-

sibility that the company will leverage the market power accruing from the

monopoly part of the business into those that are competitive. In addition,

he suggests that the pace of technical innovation will slow under the NetCo

option as service providers would have to convince NetCo that suffi cient

demand existed for it to warrant making the necessary investment.

Implementing functional separation in fi xed markets 235

Due to the problems and uncertainties associated with both the LoopCo

and NetCo proposals, it is not surprising that attention has focused on

other ways of resolving the tensions that exist within the telecommunica-

tions market. Cave (2006a, p. 94), for example, identifi es six alternatives

as follows:

creation of a wholesale division; ●

virtual separation; ●

business separation; ●

business separation with localized incentives; ●

business separation with separate governance arrangements; ●

legal separation. ●

According to Cave (2006a, p. 94), accounting separation at the time

of his writing was more or less the modus operandi of European incum-

bents with the exception of BT. In terms of the six identifi ed alternatives,

BT–Openreach falls under business separation with local incentives and

consequently is closer to the ownership separation end than the account-

ing end. BT–Openreach is an example of functional separation. According

to the European Regulators Group (ERG, 2007, p. 2), this involves the

selective separation of those parts of the network that are diffi cult for

other operators to replicate but which they need to access in order to

provide their own services. Although this is frequently interpreted as the

separation of the incumbent’s wholesale and retail businesses from one

another, this is slightly misleading. If the focus is solely on those parts of

the incumbent’s network that cannot be replicated then the scope of the

functional separation may be narrower than is implied by the separation

of wholesale from retail (ERG, 2007, p. 8).

Regardless of the extent to which functional separation is implemented,

the result is to run and manage one part of the network separately from the

rest. The separated part of the incumbent should be provided with local

incentives so that it acts in the interests of all its customers, internal and

external, and not in the interest of its parent company.2 In addition to the

use of local incentives, the ERG (2007, pp. 2f) suggests a range of ‘key ele-

ments’ that need to be provided if functional separation is to be eff ective.3

11.3 BACKGROUND

Before providing a brief overview of the events that culminated in the

establishment of Openreach in late 2005, it is necessary to note that the

decision to initiate the strategic review of telecommunication can be


located at the confl uence of three sets of drivers. The fi rst of these drivers

was the need to incorporate EU directives into the UK regulatory frame-

work, while the second was the relatively recent establishment of Ofcom

in 2003. Thus, the strategic review could be viewed as drawing a line under

the old regulatory framework and providing a basis on which converged

regulation could progress.

The review could also be regarded as being a reponse to a third set of

drivers, namely, the failure of competition to develop as anticipated in the

UK. Although some companies had invested in their own infrastructure,

these networks lacked scale (Ofcom, 2004a, p. 53). The cable operators,

which operated the most extensive networks geographically, collectively

covered less than half of the population and their ability to compete was

limited by their continued fi nancial woes.4 Service- based competition had

been possible since the late 1990s (Ofcom, 2004c, p. 53) but had enjoyed

only limited success because, it was alleged, BT had abused its dominant

position in the wholesale market to enhance its retail competitiveness

(Wilsdon and Jones, 2002).

The results of such anticompetitive behaviour can be seen with respect

to broadband and local loop unbundling. Although towards the end of

the 1990s many companies expressed an interest in off ering broadband

services, most subsequently left the market (Turner, 2003, p. 6). One con-

sequence of this was that only a handful of companies emerged to compete

against BT, while another was the limited uptake of local loop unbundling

in the UK.5 With this in mind, the strategic review could also be viewed as

being driven by the desire to enhance competition within the broadband

telecommunications market and to encourage greater adoption of local

loop unbundling.

11.3.1 The Strategic Review of Telecommunications

At the end of 2003, Ofcom announced its intention to hold a review

of the telecommunications market during the following year (Ofcom,

2004a). The initial consultation document was wide- ranging in nature,

raising issues that were subsequently clarifi ed in the second consultation

document (Ofcom, 2004b). Central to the second consultation document

was the identifi cation of three regulatory options, the fi rst of which was

deregulation. Ofcom concluded, however, that this was not possible, not

least because sector- specifi c regulation was faster and more precise than

the alternatives.

The second option was a reference under the Enterprise Act (2002) to

the Competition Commission. Such a reference would inevitably neces-

sitate a wide- ranging review of the telecommunications market that could


result in the eventual imposition of structural remedies. The third option,

the one preferred by Ofcom, was termed ‘real equality of access’ and

would enable those companies purchasing wholesale products from BT to

do so on the same terms as BT’s own retail operations. Thus, wholesale

customers would have access to (Ofcom, 2004c, p. 14):

the same or a similar set of regulated wholesale products as BT’s ●

own retail activities;

at the same prices as ● BT’s own retail activities; and

using the same or similar transactional processes as ● BT’s own retail

activities.

Two diff erent types of equivalence were proposed, outcome and input,

and a range of products identifi ed where it could be applied (Ofcom,

2004b, p. 68). In the case of equivalence of outcome, wholesale customers

receive products that are comparable to those off ered to BT’s own retail

operations but the underlying processes would not be the same. In con-

trast, where equivalence of input is applied, wholesale customers receive

the same products as BT’s own retail operations using the same set of

underlying processes (Ofcom, 2004c, pp. 67f).

Recognizing that a range of issues had been identifi ed by many of BT’s

wholesale customers during the consultative process, which these custom-

ers believed placed them at a competitive disadvantage relative to BT (ibid.,

p. 70), there was also a behavioural dimension to equivalence. While the

range of issues highlighted was broad, two areas in particular – namely the

incentives for inappropriate behaviour and transparency – were singled

out as areas where action could be taken. This said, Ofcom did note that

BT had, in the past, devoted considerable eff ort and resources to address-

ing the complaints raised by its competitors.

11.3.2 Undertakings in Lieu of a Reference under the Enterprise Act 2002

In June 2005, Ofcom (2005a) announced that it was launching a consulta-

tion to determine whether it should accept the undertakings off ered by

BT to bring an end to the strategic review. Rather than trigger a reference

under the Enterprise Act 2002, BT agreed to a series of legally enforceable

undertakings (ibid., p. 2). BT agreed to create an access service division

that would (Odell, 2005, p. 23):

control the ‘last mile’ of the telecommunications network; ●

be operationally independent of ● BT while remaining under its

ownership;


be branded diff erently from BT; ●

have its own fi ve- member board, headed by a non- executive director ●

of BT;

incorporate 15 000 of BT ● Wholesale’s 28 000 employees.

In addition, BT also agreed to a schedule for equivalence for legacy

products as well as stating the principles on which the company’s next

generation network (NGN) would be developed (Ofcom, 2005a, pp. 2ff ).

For its part, Ofcom stated that it would revisit issues such as leased lines

and retail price controls in the near future (ibid., p. 5).

In September 2005, Ofcom accepted the undertakings off ered by BT

(Ofcom, 2005b). In total, 236 undertakings were made by BT. These

governed the operation of the access service division to ensure that those

wholesale customers reliant upon access to deliver their own products

and services were treated no diff erently from BT’s own retail operations

(Ofcom, 2005c). At the same time as Ofcom agreed to accept the under-

takings, BT rebranded its access service division as Openreach (Ofcom,

2005b).

11.4 IMPLEMENTING THE UNDERTAKINGS

Given the magnitude of the undertakings, it is no surprise that their imple-

mentation has been carefully monitored. During 2006 and 2007 Ofcom

published two evaluations of the impact of the telecommunications strate-

gic review that detail the progress that BT has made in implementing the

undertakings. In addition, fi ve quarterly reviews as well as correspondence

between Ofcom, BT and others have been published.6

A useful starting point for an understanding of how the undertakings

have been implemented is the two annual evaluations that have been pub-

lished by Ofcom. The fi rst of these, which was published in October 2006,

acknowledged the eff ort that BT had invested into meeting the undertak-

ings before identifying a range of areas where implementation had been

less than satisfactory (Ofcom, 2006a). Eight areas where further action was

required were identifi ed. It is, perhaps, no surprise that these were broad

in their scope, ranging from the need to resolve boundary issues between

BT Wholesale and Openreach to agreeing how Openreach’s management

information systems (MIS) and operational support systems (OSS) could

be separated out from the rest of BT.7

The second annual report suggested that further work was required to

separate Openreach from the rest of BT and to develop, and subsequently

deploy, equivalent products (Ofcom, 2007b). In addition, the report also


stated (on p. 4), somewhat vaguely, that more eff ort was required if the full

benefi ts of functional separation were to be achieved.

Both annual reports highlighted the diffi culties to be faced in separat-

ing Openreach from the rest of BT, noting in particular the information

system- based diffi culties being encountered. The three information systems

in question are the MIS, OSS and the equivalence management platform

(EMP). Openreach is required to separate its OSS from the rest of BT in a

logical manner and to separate them physically from one another by June

2010 (Ofcom, 2007b). However, Ofcom and BT placed a diff erent interpre-

tation on logical separation, with the consequence that clarifi cation was

required (Ofcom, 2007d). Although this inevitably resulted in some delays,

it also produced a clear timetable for the migration of users to physically

separate systems.

Additional time was also sought by BT to separate the MIS between

Openreach and the rest of the company. While Ofcom did agree to this

request, BT was required to assist users to restrict access and to defi ne the

subsequent separation process. This has largely been achieved although

Ofcom does note (2007a, p. 48) that risks still remain with those systems

that draw on BT- wide initiatives. The delivery of equivalence is supported

through the use of the EMP,8 the implementation of which has been less

than satisfactory since it was fi rst introduced in early 2006. The delivery of

the initial system was delayed and subsequent versions were released with

reduced functionality (OTA, 2006a, 2006b). Perhaps more importantly,

concerns have been raised as to the stability of the EMP (OTA, 2007a) and

the extent to which the service is unavailable (OTA, 2007b, 2007c). Both of

these have caused problems for the telecommunications companies using

the EMP. Although these issues have been tackled with varying degrees of

success, they have engendered a degree of uncertainty regarding the plat-

form’s robustness and reliability.

The correspondence published by Ofcom highlights some of the diffi cul-

ties that have been encountered in the implementation of the undertak-

ings.9 BT has, on more than one occasion, sought more time to implement

the undertakings. Although Ofcom has invariably granted these requests,

it is worth noting that the extensions are temporary and not open- ended.

This correspondence is relatively brief, which is in contrast to the consul-

tations surrounding the range of exemptions and variations that BT has

sought since June 2006.

To bring the strategic review to a swift conclusion, it was agreed that

Ofcom and BT could consider at a later date the equivalence of inputs

needs of some products (Ofcom, 2006e). In June 2006, BT sought exemp-

tions and variations in 15 diff erent areas, with 13 requiring consultation

due to their complexity (Ofcom, 2006f). As a result of this consultation,


Ofcom agreed to nine of the requests. The remaining four requests

required additional consultation as they involved products relying on fi ber

(Ofcom, 2006e).

In turn, this additional consultation resulted in three out of the four

requests being granted. Ofcom granted a temporary extension until

December 2007 in the case of the fourth request while further consultation

was undertaken (Ofcom, 2007e, p. 2). Another set of exemptions and vari-

ations granted to BT was published in October 2007 and was once again

the outcome of a consultation process that began in July of the same year

(Ofcom, 2007f). The variation request made in May 2008 granted BT and

Openreach more time to implement the undertakings (Ofcom, 2008b). All

18 granted variations are shown in Table 11.1.

The published correspondence, as well as the exemptions and variations

consultations published by Ofcom, draw attention to boundary issues.

Boundary issues arise where the distinction between Openreach and the

rest of BT is blurred. One area where boundary issues have arisen was

noted above, namely to ensure that the information systems that linked

Openreach with the rest of BT were altered so that the two were separate

from one another.

A second area where boundary issues have emerged is that of access

to engineering resources. BT has sought to move engineers between

Openreach and BT Wholesale as circumstances dictate (Ofcom, 2006b, p.

3). For example, BT requested permission from Ofcom to move engineers

between the two divisions in the aftermath of the fl oods that swept the

south of England during 2007. This was, however, a temporary measure

that addressed a particular series of events.

The quarterly reports published by Ofcom draw attention to the need to

ensure that the ‘Chinese Walls’ between Openreach and the rest of BT are

maintained (Ofcom, 2006b, 2006d). A separate Openreach head offi ce has

been established, and some of the earlier concerns that the ‘Chinese walls’

were unsatisfactory due to organizational changes within BT Wholesale

have been addressed.10 The second report on the implementation of the

undertakings notes that Openreach is reliant on other parts of BT for

access to space and power within exchanges (Ofcom, 2007b, pp. 46ff ). As

a consequence of this, it was felt that Openreach does not have adequate

control over the products that it delivers.

Also highlighted by the quarterly reports are the concerns expressed by

other telecommunication operators as well as by Ofcom regarding product

development. The former have noted that a gap existed early on between

what BT announced and what was delivered (Ofcom, 2006b, p. 11), while

more recently their interaction with BT has become an issue (Ofcom,

2006c, pp. 9ff ; Ofcom, 2006d, p. 10). It was alleged that this interaction


was insuffi cient, and that in some cases the ability of other telecommuni-

cation operators to infl uence product specifi cations was limited. To this,

Ofcom (2007b) adds that the pace of product development has been slow,

before acknowledging that this may be due to developments elsewhere in

BT.11

Table 11.1 Variations to the undertakings

Variation Scope Date

1 Products & services supplied by Access

Services (Openreach); share schemes and BT

Group Deferred Bonus Plan; EAB report to

OFCOM; EAB Summary Annual Report

March 2006

2 Equipment location April 2006


Services (Openreach)

August 2006

4 OSS separation September 2006

5 OSS separation October 2006



December 2006

7 Information fl ows & system separation December 2006



April 2007

9 OSS separation June 2007

10 Incident management processes October 2007

11 Extensions to OSS and EOI timetables November 2007

12 Changes to sections 2.1 (defi nitions), 5 (access

services) and 6 (management & structure of BT

Wholesale)

December 2007



December 2007

14 Provision of equivalent products and services

– changes to section 3.1.1, section 3.1.2 and

annex 1 of the undertakings

December 2007


Services, changes to section 5.46.2

May 2008

16 NGN, Space and Power and OSS separation October 2008

17 IPStream in certain geographic markets and

Wavestream National

December 2008

18 Changes to sections 2.1 (defi nitions) and 5.46.2

(date of eff ect of undertakings)

December 2008

Source: www.ofcom.org.uk/telecoms/btundertakings/exemptionsandvariations, accessed 1 June 2009.


11.5 THE IMPACT OF FUNCTIONAL SEPARATION

The previous section has demonstrated that the separation of Openreach

from the rest of BT has not been straightforward. While Openreach was

established relatively swiftly in the aftermath of the undertakings being

agreed, the actual separation of Openreach and BT has been more prob-

lematic in terms of both products and processes. As the implementation

of the undertakings has progressed, BT has sought exemptions, variations

and extensions. Such requests were perhaps inevitable given the unprece-

dented nature of the undertakings and thus may simply refl ect the inherent

diffi culties of separating Openreach from the rest of BT.

Notwithstanding the diffi culties that have been experienced, the

Openreach model of functional separation has been praised. In a speech

to the ERG, Commissioner Reding argued that functional separation had

contributed to the rapid rise of unbundled lines and increased network

investment (Reding, 2007). When the undertakings were announced in

September 2005, the number of unbundled lines stood at 123 000 (OTA,

2005). In contrast, the most recent fi gures published in May 2009 show

that there are now 5.9 million unbundled lines (OTA, 2009). Although

this growth, which is shown in Figure 11.1, is undoubtedly impressive,

it has not been without its problems as we have shown above. Monthly

updates from the Offi ce of the Telecommunications Adjudicator (OTA)

show that although progress has been made in implementing the under-

takings, with many key performance indicators improving after initial

7000

6000

5000

4000

3000

2000

1000

0

Date

000 L

LU

Sep-0

5

Nov-

05

Mar

-06

May

-06

Jul-0

6

Sep-0

6

Nov-

06

Jan-0

7

Mar

-07

May

-07

Jul-0

7

Sep-0

7

Nov-

07

Jan-0

8

Mar

-08

May

-08

Jul-0

8

Sep-0

8

Nov-

08

Jan-0

9

Mar

-09

May

-09

Jan-0

6

Source: Monthly OTA updates, available at www.off ta.org.uk.

Figure 11.1 Unbundled local loops, September 2005–May 2009


disappointments, problems continue to emerge such as the delays in pub-

lishing the Wholesale Line Rental 3 (WLR3) roadmap and providing the

necessary functionality to other service providers (OTA, 2008a, 2008b,

2008c).

There can be no doubt that BT has invested to deliver the undertakings.

According to its 2008 annual report, BT invested £35 million in a ‘proac-

tive maintenance programme’ that reduced the number of faults experi-

enced (BT, 2008, p. 21). The total amount invested by BT in the fi nancial

year ending 31 March 2008 was £53 million, which was an increase on

the corresponding fi gure for 2007 (£30 million) but less than in 2006 (£70

million) (ibid., p. 105). This is, however, slightly misleading as these fi gures

relate to the costs of establishing Openreach and meeting the undertak-

ings.12 In other words, they do not include capital expenditure. Over the

same period, capital expenditure remained steady at more or less £1.1

billion per annum.13 It is not clear, of course, how much of this would have

occurred regardless of whether Openreach was established, and how much

is specifi cally due to the implementation of the undertakings.

There are other service providers in the marketplace. One such opera-

tor is Cable & Wireless (C&W). Prior to the undertakings being agreed,

C&W stated its intention to invest in local loop unbundling (LLU)

(Cable & Wireless, 2004) and acquired Bulldog, a broadband provider, to

strengthen its position in the marketplace. After failing to control costs,

C&W announced in June 2006 that it would stop off ering retail products

(Staff ord, 2006) although it has continued to use LLU to deliver services

to business clients (Cable & Wireless, 2008).14

Another operator providing broadband services is BSkyB. As of 31

March 2008, BSkyB had 1.428 million broadband subscribers, an increase

of 229 000 subscribers over the quarter (BSkyB, 2008). In addition to acquir-

ing Easynet at the end of 2005 for £211 million (Wray and Milmo, 2005),

BSkyB has also made subsequent investments in infrastructure. In the nine

months ending 30 April 2008, BSkyB invested £127 million in its residential

broadband and telephony business (BSkyB, 2008). While it is not clear how

much of this investment was targeted towards broadband, it is worth noting

that it has been claimed that this represents the peak of its broadband-

related investment cycle (Edgecliff e- Johnson and Fenton, 2008).

A third operator in the marketplace is Carphone Warehouse. Since

announcing its intention to use LLU in November 2005, Carphone

Warehouse has grown to become a signifi cant player in the market. The

economics of LLU will be further enhanced when Carphone Warehouse

completes its £236 million acquisition of the UK operations of Tiscali

that was announced in May 2009 (BBC, 2009). The company’s growth

has been driven by innovative new products such as TalkTalk Free


Broadband, launched in April 2006, which gave free broadband to those

customers purchasing voice and line rental (Carphone Warehouse, 2007,

p. 6). Although this was highly successful, attracting more than 500 000

customers, losses associated with the product widened from £20 million

in April to £70 million in October 2006 (Parker and Braithwaite, 2006).

Accompanying the launch of TalkTalk Free Broadband was the inten-

tion to invest in 1000 unbundled exchanges by July 2007 (Carphone

Warehouse, 2007, p. 6). One motive for investing in LLU was that deliver-

ing services to customers this way was profi table, unlike the situation when

wholesale products were used (Pratley, 2006).15

The acquisition of AOL UK from Time Warner in December 2006

further expanded the company’s customer base.16 This not only served

to increase the number of subscribers but would make the economics of

LLU more attractive to Carphone Warehouse (Parker and Braithwaite,

2006, p. 23). As a consequence, it is perhaps no surprise that Carphone

Warehouse has switched the bulk of its subscriber base onto LLU and

expanded the number of exchanges in which it has invested (Carphone

Warehouse, 2008, p. 5).17 Although the expansion of the business would

not be possible without infrastructure investment of one sort or another,

it is not clear how much the company has actually invested as the relevant

fi gures are consolidated with other investments in its annual report.18 This

said, Charles Dunstone, the Chief Executive of Carphone Warehouse, has

recently been quoted as saying that the company has made large expendi-

ture commitments towards unbundling (Parker, 2008a, p. 19).

Dunstone has also raised the issue of the fees that Carphone Warehouse

pays Openreach. Charge ceilings for WLR and LLU services were set by

Ofcom between December 2004 and January 2006.19 Since then, Openreach

has failed to achieve the 10 percent return that it is permitted, with the con-

sequence that BT would like to raise these charges (Parker, 2008e, p. 18).

BT has argued that the aforementioned charge ceilings do not refl ect the

underlying costs of providing services and that they need to be amended

to refl ect the changing nature of the product portfolio being delivered.20

Unsurprisingly Carphone Warehouse has a diff erent view, stating that it

would unfair to ‘change the game’ (Parker, 2008a, p. 19).

Notwithstanding the complexities of determining rates of return and

charges that Ofcom (2008a, 2008d) highlights, the implications are clear:

if Ofcom agreed with BT, the costs for other service providers of using

Openreach’s network would increase. In mid- May 2009 Ofcom announced

that Openreach would be allowed to increase the prices that it charges

other operators for both metallic path facilities and shared metallic path

facilities (Ofcom, 2009).21 The price of metallic path facilities would rise to

£86.50 and that for shared metallic path facilities to £15.80 (Ofcom, 2009,


p. 4). Signifi cantly these increases were towards the lower end of the range

suggested in the consultation document (Ofcom, 2008d), leading BT to

state subsequently that it may challenge this decision (Parker, 2009).22

BT is also quoted as noting that the lower charges were a ‘real disincen-

tive for future investments’ (Parker, 2009). In its 2009 annual report, BT

states that the regulatory environment must encourage investment and

provide a rate of return commensurate with the risk involved (BT, 2009).

Whilst BT does not say what constitutes a reasonable rate of return, it

does note that regulatory clarity and certainty are important factors when

determining whether the company will invest or not. It is arguably the case

that Ofcom has provided regulatory stability through outlining how prices

will increase in the coming years, and clarity through extensively explain-

ing how it has arrived at its conclusions. Although Ofcom (2009, p. 30)

does acknowledge that the price changes will have an impact on telecom-

munication companies, it then goes on to state that it is more interested in

providing a stable and predictable regulatory framework than guarantee-

ing the returns of individual companies within the market.

Within such an environment, the inevitable question is: Why has BT

continued to invest in Openreach? One reason is that Openreach is profi t-

able and that the investments that have been made are intended to main-

tain this profi tability.23 In other words, it is in BT’s economic interest to

invest in Openreach. Thus, the complaints that BT has made regarding

the rate of return that it is allowed to make on such investments could be

interpreted as being part of a ‘game’ that is played between Ofcom, other

telecommunication companies and itself. Another explanation could be

that notwithstanding the undertakings that have been implemented, BT

still derives suffi cient advantages from owning the last mile of the network

to warrant retaining ownership of, and investing in, Openreach. It may

also be the case that BT has continued to invest in Openreach because it

believes that through doing so it will be treated more favorably in other

markets.

In the process of assessing the costs and benefi ts of functional separa-

tion, Amendola et al. (2007) state that functional separation enhances

competition. One measure of this is market share while another is the

price and speed of broadband products. As can be seen from Figure 11.2,

BT remains the largest Internet Service Provider (ISP) whereas the market

shares of Orange, Tiscali, Virgin Media and smaller ISPs have all fallen. In

contrast, the market shares of Carphone Warehouse and BSkyB have both

risen since the undertakings were agreed.

Ofcom (2008a, p. 14) states that broadband prices have fallen whereas

speeds have increased. While there is clear evidence that prices have

fallen,24 there has been much discussion of late regarding broadband


speeds. Average download speeds vary considerably across the UK, from

4.5 Mbps in London to 2.3 Mbps in Northern Ireland (BBC, 2008). In

addition, the relationship between the advertised and the actual broad-

band speeds has been questioned, with a discrepancy being noted between

the two. According to Point Topic (2008), as advertised speeds increase the

proportion of subscribers receiving such speeds declines. The discrepancy

between advertised and actual broadband speeds has prompted Ofcom

(2008c) to issue a voluntary code of conduct containing eight principles.

11.6 CONCLUSION

This chapter has focused on functional separation in the UK. The

September 2005 decision to opt for functional separation represents a

milestone in the regulation of the UK telecommunications market, and

in the process imposed a range of undertakings on BT. The process of

unraveling the numerous undertakings that bound Openreach to the rest

of BT has proved to be more diffi cult and protracted than anticipated. Not

only were extensions, variations and exemptions sought, but Openreach

also took longer than anticipated to meet several key performance indica-

tors. If nothing else, these diffi culties suggest that implementing functional

separation is by no means straightforward.

Nevertheless, the period subsequent to the adoption of functional sepa-

ration has seen signifi cant broadband competitors emerge in the form of

Carphone Warehouse and, to a lesser extent, BSkyB. For both Carphone

Warehouse and BSkyB, LLU plays a central role in their strategies, not

35

30

25

20%

15

10

5

0

Q4 2

004

Q1 2

005

Q2 2

005

Q3 2

005

Q4 2

005

Q1 2

006

Q2 2

006

Q3 2

006

Q4 2

006

Q1 2

007

Q2 2

007

Q3 2

007

Q4 2

007

Q1 2

008

Q2 2

008

Q3 2

008

BT

Carphone Warehouse

Orange

Sky

Tiscali

Virgin Media

Smaller player totals

Source: www.broadbanduk.org.

Figure 11.2 Market share of ISPs


least because LLU is a more profi table way to service their customers than

WLR. While LLU does not appear to have spurred other service provid-

ers into investing in other parts of the ‘ladder of investment’ as suggested

by Cave (2006b), they do appear to have invested relatively large sums

to deliver LLU- underpinned broadband services. With the emergence of

LLU as a vehicle for the deployment of broadband services, not only has

the importance of BT in the marketplace been elevated, but so too has the

necessity of ensuring that the relationship between BT and those compa-

nies using its network is functioning as intended.

Functional separation was adopted to resolve the tensions that existed

between BT and those companies wishing to access its network to provide

their own services, with the undertakings providing a framework for

assessing the state of this relationship. Looking back over the period since

the undertakings, it could be argued that the relationship between BT

and other service providers has improved as the number of unbundled

lines has substantially increased. However, the relationship could sour if

BT continues to argue that the rate of return that it is allowed to earn is

unsatisfactory. If the relationship between BT and other operators does

sour and undertakings are breached, Ofcom could refer the matter to the

Competition Commission. The unpredictable outcome of such a refer-

ral may mean that Ofcom is unwilling to make such a reference and will

adopt a more pragmatic approach to implementing and monitoring the

undertakings.

NOTES

1. This is not to suggest that this is the only feasible categorization of structural separa-tion. Dounoukos and Henderson (2003, pp. 44f), for example, distinguish between ‘actual’ structural separation and ‘internal’ separation. With respect to the former, they identify four alternatives – club or joint ownership, operational separation, separation into several vertically integrated companies and separation of the non- competitive com-ponents into several parts; while three ‘internal’ separation alternatives – accounting, functional and corporate – are suggested. A discussion of structural alternatives based on a broader range of regulated industries than just telecommunications can be found in OECD (2001).

2. While referring to wholesale and retail, the OECD (2006) states that incentives may be given to wholesale managers that confl ict with those provided to retail managers. This reinforces the suggestion that incentives should be localized and not tied to the overall profi tability of the operator.

3. The European Regulators Group (2007) provides a summary of the measures under-taken in countries where functional separation has been implemented under the three headings of functional, employee and information separation. Interestingly, not all measures have been implemented in all cases, and some are only feasible in conjunction with others.

4. In 2004 just 46 percent of UK homes were passed by broadband- enabled cable (Ofcom,


2004c, p. 38). Subsequent communication market reviews have shown that in 2005 and 2006, digital cable was available to 45 percent of the UK population (Ofcom, 2007a, p. 15). For a discussion of the fi nancial woes of cable operators see, for example, Curwen (2004).

5. See, for example, de Bijl and Peitz (2005) for a discussion of unbundling that highlights the relatively slow uptake of local loop unbundling in the UK compared to other European Union member states.

6. For details of the issues raised by the quarterly reports, see Ofcom (2005a, 2006b, 2005d, 2007c). These reports, as well as the other material published by Ofcom relating to the telecommunications strategic review and the implementation of the undertak-ings, can be found at www.Ofcom.org.uk/telecoms/btundertakings.

7. See Ofcom (2006a, p. 2) for a full list of the eight areas identifi ed that require further action.

8. The EMP is an information system that supports the delivery of the products off ered by Openreach. Given the anticipated large volumes of some of these products, the system is designed to be automated to ensure that services are provided as demanded.

9. See www.Ofcom.org.uk/telecoms/btundertakings/exemptionsandvariations for a full list of the correspondence between BT and Ofcom that has been published.

10. BT Wholesale has been reorganized with two management units – BT Wholesale Core Network Services and BT Wholesale Value- added Network Services – being established (Ofcom, 2006b). The concern expressed by some telecommunications operators was that this would complicate the implementation of the ‘Chinese walls’ that were estab-lished. However, Ofcom stated that more time should be given before making a judg-ment as to whether the ‘Chinese walls’ were being breached.

11. Two developments are noted, namely, the implementation of the undertakings and the development of the company’s twenty- fi rst century network.

12. It is worth noting that the annual report states that these fi gures are estimates of the incremental and directly attributable costs incurred as a consequence of establishing Openreach on the one hand and meeting the undertakings on the other (BT, 2008).

13. For the year ended 31 March 2007, capital expenditure was £1108 million while the fi gure for the following year, albeit an estimate on the part of BT, was £1100 million (Ofcom, 2008a, p. 20). Although capital expenditure has remained more or less steady since Openreach was established, the 2007 fi gure did represent a slight increase on the previous year’s level of £1038 million (Ofcom, 2008a, p. 20).

14. Cable & Wireless does not break up capital expenditure by its European, Asian and US businesses by geography or product. Thus, it is unclear how much the company has invested in LLU in the UK. Having said this, capital expenditure by this business did decrease from £235 million in 2006/07 to £221 million in 2007/08 (Cable & Wireless, 2008).

15. According to Pratley (2006), Carphone Warehouse would make a profi t of £7 per month per customer when using LLU whilst it would lose £5 per month when using wholesale products.

16. With the acquisition of AOL UK, Carphone Warehouse controlled 16 percent of the broadband market (Carphone Warehouse, 2007). While not without its risks, the acquisition has also been regarded as strategic in nature as it increased the presence of Carphone Warehouse in this market (Parker, 2008b, 2008c). Carphone Warehouse sought to expand its presence in the UK broadband market by bidding to acquire the UK operations of Tiscali, which had 1.8 million broadband subscribers at the time. Although Parker (2008d) reports that C&W is no longer in contention, other reports published around the same time contest this.

17. At the end of 2006, 19 percent of the company’s broadband subscribers were located in unbundled exchanges (Carphone Warehouse, 2008). This fi gure had increased to 61 percent by the end of the following year. Over the same period, the number of unbun-dled exchanges increased from 569 to 2457. Only part of this increase – 924 – was due to the acquisition of AOL UK.


18. Total investment increased from £351.1 million in 2006 to £562.2 million in 2007 (Carphone Warehouse, 2007), with ‘acquisition of property, plant and equipment’ increasing from £89.4 million to £161.4 million over the same period (Carphone Warehouse, 2007). Some, but not all, of this presumably relates to LLU.

19. See Ofcom (2008a, pp. 1f) for details. The charge ceilings were as follows: £100.68 for residential WLR, £110.00 for business WLR, £81.69 for metallic path facilities (MPF) and £15.60 for shared mpf (SMPF).

20. The move away from WLR has, for example, contributed to a decline in Openreach’s rate of return (Ofcom, 2008a, p. 18).

21. See, for instance, Ofcom (2009), for a detailed description of the new pricing arrange-ments for Openreach.

22. As part of this decision to set prices at the lower end of the range suggested in the con-sultation document, Ofcom declined to allow BT to use the Openreach’s fees to reduce BT’s pension defi cit (Parker, 2009). For an assessment of BT’s pension defi cit see, for instance, Ralfe (2009).

23. Openreach is one of BT’s four operating divisions. According to its 2009 annual report, three of these divisions – BT Retail, BT Wholesale and Openreach – maintained or increased their levels of operating profi ts for the year ending 31 March 2009 compared to 2008, whilst the fourth division – BT Global Services – incurred a substantial operat-ing loss (BT, 2009). Between 2008 and 2009, Openreach increased its operating profi t from £1050 million to £1209 million whilst BT Global Services saw its 2008 operating profi t of £117 million turn into a loss of £2106 million in 2009.

24. See, for example, Ofcom (2006g, p. 120), Ofcom (2007a, p. 298) or Ofcom (2008a, p. 15).

REFERENCES

Amendola, G., F. Castelli and P. Serdengecti (2007). Is really functional separation the next milestone in telecommunications (de)regulation? Paper presented at the 18th European Regional ITS Conference, Istanbul, Turkey, 2–5 September.

BBC (2008). Towns triumph in broadband tests. 3 June. Available at news.bbc.co.uk.

BBC (2009). Carphone to purchase Tiscali UK. 8 May. Available at news.bbc.co.uk.

de Bijl, P. and M. Pietz (2005). Local loop unbundling in Europe: experience, pros-pects and policy challenges. Communications and Strategies, 57, 33–57.

BSkyB (2008). 3rd quarter results. 30 April. Available at www.bskyb.com.BT (2008). Keeping BT ahead of the game. Annual Report & Form 20- F.

Available at www.bt.com.BT (2009). Annual Report & Form 20- F. Available at www.bt.com.Cable & Wireless (2004) Press release – Cable and Wireless PLC presentation on

local loop unbundling. 16 September. Available at www.cwplc.com.Cable & Wireless (2008). Results for the year ended 31 March 2008. 22 May.

Available at www.cwplc.com.Carphone Warehouse Group (2007). Annual Report 2007 – Delivering value to

our customers and shareholders. Available at www.cpwplc.com.Carphone Warehouse Group (2008). Preliminary results for the 52 weeks to 29

March 2008. Available at www.cpwplc.com.Cave, M. (2002). Is LoopCo the answer? info, 4 (4), 25–31.Cave, M. (2006a). Six degrees of separation: operational separation as a remedy


in European telecommunications regulation. Communications and Strategies, 64, 89–103.

Cave, M. (2006b). Encouraging infrastructure competition via the ladder of invest-ment. Telecommunications Policy, 30, 223–37.

Curwen, P. (2004). Markets vs regulators and the battle to determine market struc-ture: evidence from the UK cable industry. info, 6 (1), 24–36.

Dounoukos, S. and A. Henderson (2003). Structural separation in telecommunica-tions: a review of some issues. Agenda, 10 (1), 43–60.

Edgecliff e- Johnson, A. and B. Fenton (2008). BSkyB hits peak of its investment in broadband. Financial Times, 7 February, 20.

European Regulators Group (ERG) (2007). ERG opinion on functional separa-tion. ERG (07) 44. Available at www.erg.eu.int.

Odell, M. (2005). New unit emerges from the shake- up. Financial Times, 24 June, p. 23.

OECD (2001). Structural separation in regulated industries. DAFFE/CLP(2001)11, Paris: OECD.

OECD (2006). Report to the Council on experiences on the implementation of the recommendation concerning structural separation in regulated industries. C(2006)650, Paris: OECD.

Ofcom (2004a). Strategic review of telecommunications – phase 1 consultation document. April, London: Ofcom.

Ofcom (2004b). The Communications Market 2004 – Telecommunications. August, London: Ofcom.

Ofcom (2004c). Strategic review of telecommunications – phase 2 consultation document. 18 November. London: Ofcom.

Ofcom (2005a). Telecommunications statement. June. London: Ofcom.Ofcom (2005b). Ofcom accepts undertakings from Board of BT Group on opera-

tional separation. BT access service business – Openreach – formally established today. 22 September. London: Ofcom.

Ofcom (2005c). Final statement on the strategic review of telecommunications and undertakings in lieu of a reference under the Enterprise Act 2002 – Statement. 22 September. London: Ofcom.

Ofcom (2005d). Report on the implementation of BT undertakings – fi rst quarterly report. 25 October. London: Ofcom.

Ofcom (2006a). Evaluating the impact of the Telecoms Review – an interim report one year on. 18 October. London: Ofcom.

Ofcom (2006b). Report on the implementation of BT undertakings – second quarterly report. 26 January. London: Ofcom.

Ofcom (2006c). Report on the implementation of BT undertakings – third quarterly report. 23 May. London: Ofcom.

Ofcom (2006d). Report on the implementation of BT undertakings – fourth quarterly report. 5 September. London: Ofcom.

Ofcom (2006e). Requests from BT for specifi ed exemptions and agreements to its Undertakings under the Enterprise Act 2002 – Part 2. 21 December. London: Ofcom.

Ofcom (2006f). Requests from BT for specifi ed exemptions and agreements to its Undertakings under the Enterprise Act 2002 – Part 1. 15 June. London: Ofcom.

Ofcom (2006g). The communications market. 10 August. London: Ofcom.Ofcom (2007a). Communications market review. 23 August. London: Ofcom.


Ofcom (2007b). Impact of the Telecoms Strategic Review. Evaluation. 10 December. London: Ofcom.

Ofcom (2007c). Report on the implementation of BT undertakings – fi fth quarterly report. 12 February. London: Ofcom.

Ofcom (2007d). BT OSS separation. 21 June. London: Ofcom.Ofcom (2007e). Requests from BT for specifi ed exemptions and agreements to

its Undertakings under the Enterprise Act 2002 – Part 3. 19 July. London: Ofcom.

Ofcom (2007f). Requests from BT for specifi ed exemptions and agreements to its Undertakings under the Enterprise Act 2002 – Part 4. 3 October. London: Ofcom.

Ofcom (2008a). A new pricing framework for Openreach. Developing new charge controls for wholesale line rental. unbundled local loops and related services. 30 May. London: Ofcom.

Ofcom (2008b). Variation of the undertakings given to Ofcom by BT pursuant to the Enterprise Act 2002 – variation number: 15. 20 May. London: Ofcom.

Ofcom (2008c). Voluntary code of practice: broadband speeds. 5 June. London: Ofcom.

Ofcom (2008d). A new pricing framework for Openreach. 5 December. London: Ofcom.

Ofcom (2009). A new pricing framework for Openreach. Statement. 22 May. London: Ofcom.

OTA (2005). Update for September 2005. Available at www.off ta.org.uk.OTA (2006a). Update for March 2006. Available at www.off ta.org.uk.OTA (2006b). Update for April 2006. Available at www.off ta.org.uk.OTA (2007a). Update for January 2007. Available at www.off ta.org.uk.OTA (2007b). Update for April 2007. Available at www.off ta.org.uk.OTA (2007c). Update for June 2007. Available at www.off ta.org.uk.OTA (2008a). Update for February 2008. Available at www.off ta.org.uk.OTA (2008b). Update for March 2008. Available at www.off ta.org.uk.OTA (2008c). Update for April 2008. Available at www.off ta.org.uk.OTA (2009). Update for May 2009. Available at www.off ta.org.uk.Parker, A. (2008a). Carphone hits at BT plans to raise line fees. Financial Times,

14 April, p. 19.Parker, A. (2008b). Risky move into broadband looks as if it might pay off .

Financial Times, 14 April, p. 19.Parker, A. (2008c). Carphone deal opens up lines of expansion. Financial Times,

9 May, p. 19.Parker, A. (2008d). Carphone off Tiscali buyer list. Financial Times, 13 May, p.

17.Parker, A. (2008e). BT seeking increased landline charges. Financial Times, 16

May, p. 18.Parker, A. (2009). Regulator keeps line open for BT. Financial Times, 22 May.

Available at www.ft.com.Parker, A. and T. Braithwaite (2006). Carphone is catapulted into the big league.

Financial Times, 12 October, p. 23.Point Topic (2008). How fast is your broadband? 12 March. Available at point-

topic.com.Pratley, N. (2006). Broadband disaster should turn to success. Guardian, 12

October. Available at www.guardianunlimited.co.uk.


Ralfe, J. (2009). Behind BT’s deep, dark pension defi cit. Financial Times, 24 May. Available at www.ft.com.

Reding, V. (2007). Better regulation for a single market in telecoms, Speech/07/624, Plenary Meeting of the European Regulators Group, Athens, 11 October.

Sandbach, J. (2001). Levering open the local loop: shaping BT for broadband competition. info, 3(3), 195–202.

Staff ord, P. (2006). C&W’s Bulldog turns tail from broadband market. Financial Times, 9 June, p. 22.

Turner, C. (2003). Issues in the mass market deployment of broadband. info, 5(2), 3–7.

Wilsdon, J. and D. Jones (2002). The Politics of Bandwidth: Network Innovation and Regulation in Broadband Britain, London: Demos.

Wray, R. and D. Milmo (2005). Watchdog clears BSkyB acquisition of Easynet. Guardian, 31 December. Available at www.guardianunlimited.co.uk.

Xavier, P. and D. Ypsilanti (2004). Is the case for structural separation of the local loop persuasive? info, 6(2), 74–92.

253

12. Effi ciency and sustainability of network neutrality proposals1

Toshiya Jitsuzumi

12.1 INTRODUCTION

Recent times have witnessed the phenomenon of increasing congestion on

the Internet; this is characterized by the ‘crowding out’ of average users

by peer- to- peer (P2P) users due to a rapid spread of bit- intensive appli-

cations (Cisco Systems, 2008a, 2008b, 2008c; Swanson, 2007; Swanson

and Gilder, 2008), and further, it is sometimes accompanied by the

possibly anticompetitive conduct of network operators. In an attempt

to deal with this, ‘network neutrality’, a term coined by Wu (2003),

has come to be a focus of discussion among operators, academics and

telecom regulators. For example in parallel with a series of controversial

events, for example, blocking of Voice over Internet Protocol (VoIP) by

Madison River Communications LLC in March 2005 and the allegedly

discriminatory treatment of P2P by Comcast since 2007,2 the US Federal

Communications Commission (FCC) has issued important statements.

First, in 2004, the then FCC Chairman Michael Powell challenged the

broadband network industry to preserve the four ‘Internet Freedoms’:

freedom to access content, freedom to use applications, freedom to

attach personal devices and freedom to obtain service plan information.3

Secondly, on 23 September 2005, the following principles were incorpo-

rated into a policy statement.4

Consumers are entitled to access lawful Internet content of their ●

choice.

Consumers are entitled to run applications and use services of their ●

choice; subject to the requirements of law enforcement.

Consumers are entitled to connect their choice of legal devices that ●

do not harm the network.

Consumers are entitled to competition among network providers, ●

application and service providers, and content providers.


Moreover the FCC mentioned the necessity of preserving these principles

in the conditions agreed upon for the October 2005 approval of both the

Verizon–MCI and the SBC–AT&T mergers and for AT&T’s concessions

in its merger with BellSouth, which was fi nally approved on 29 December

2006. In addition the US Federal Trade Commission (FTC) issued a policy

report in 2007 (FTC, 2007), and several bills have been proposed and were

proposed and discussed in the 109th and 110th Congress.5

In Japan, the Ministry of Internal Aff airs and Communications (MIC)

issued a report in September 2007 regarding network neutrality (MIC,

2007b). Similar considerations are being made in the EU as well.

Since the concept of network neutrality encompasses many diff erent

aspects, a discussion on the related issues has thus far been quite complex.

Therefore, this chapter divides network neutrality issues into economic

and non- economic ones and focuses mainly on the former, which is a

direct result of traffi c congestion and negative externality thereof. I further

categorize the network neutrality challenges into short- and long- term

issues and discuss policy recommendations. Short- term solutions include

a case- by- case approach and a re- evaluation of the ‘smart market’ concept

proposed by MacKie- Mason and Varian (1998), while long- term solu-

tions involve the need for a new business model and the introduction of a

mechanism similar to a universal service fund (USF).

The remainder of this chapter is organized as follows. In section 12.2,

the defi nition of network neutrality is analyzed. Section 12.3 discusses

some basic concepts that make the network neutrality issue unique, and

section 12.4 proposes policy recommendations for general cases. Section

12.5 concludes the discussion.

12.2 MULTIPLE ASPECTS OF NETWORK NEUTRALITY

The most important factor that has contributed to the complexity of the

discussions on the topic under consideration is the lack of a standard-

ized defi nition of network neutrality. Examining the related arguments

that have been presented thus far, there are at least six dimensions to

this concept, as discussed below. However, some of them are not strictly

orthogonal from each other.

12.2.1 Economic Aspect versus Non- economic Aspect

Some discuss this issue from the viewpoint of optimal resource allocation

or maximizing economic effi ciency; others underscore the non- economic

Effi ciency and sustainability of network neutrality proposals 255

value of this ‘neutrality’. For example, network users stress the value of the

Internet as a medium and relate it with freedom of speech or democracy,

while network operators emphasize the effi cient use of network. However

it is important to remember that these two aspects are not mutually exclu-

sive. Rather the non- economic aspect works as an important factor modi-

fying the naive outcome generated by a pure economic consideration. In

fact the fundamental diff erence between network users and operators lies

in how they perceive these two aspects.

12.2.2 Long Term versus Short or Very Short Term

Faced with congestion on the Internet, some people mention the impor-

tance of network investment or technological development in alleviating

capacity constraints in the long term, while others focus on how to control

peak demand, implicitly assuming that the network capacity is fi xed in

the short term. However, as Yoo (2006) and Peha (2007) point out, these

windows are not necessarily mutually exclusive but merely diff erent sides

of the same coin.

Moreover, in the long term, prospects of future technological develop-

ments may become an important element of this discussion. It is important

to bear in mind that a short- term solution that assumes constant technol-

ogy can be mutually incompatible with a long- term solution that takes

some technological development into consideration; in other words, the

static- effi cient solution may be detrimental to the dynamic- effi cient one.

For example a local loop unbundling (LLU) to incubate access competi-

tion may have diff erent impacts on diff erent time horizons.

12.2.3 Fixed Telecom Network versus Cable Network or Mobile Network

Congestion requires government intervention only when the resources in

question are scarce and shared as ‘common goods’. Thus, theoretically

speaking, congestion can be a problem for the trunk/backhaul/backbone

portion of the network and for the access network of cable and mobile

services, while it does not pose a problem for telcos’ access segments where

certain capacities are usually reserved for individual users.6

12.2.4 Single- Sided Market versus Two- Sided Market7

On some occasions, a solution to network neutrality can be developed

assuming that subscription revenue is the only source of an operator’s

revenue. Alternatively one can assume that network operators can expect

additional revenues from content and/or application providers (two- sided


market) and can therefore utilize the advertisement model as an integral

part of a solution.

12.2.5 Unbundled Market versus Bundled Market

In reality, network operators are either vertically separated from content

and application providers or vertically integrated with them. Since network

operators usually possess a certain market power, vertical integration

requires additional considerations of leveraging, extraction, discrimina-

tion or price squeezing.

12.2.6 Legal Use versus Illegal Use

A signifi cant portion of Internet traffi c reportedly comprises unlaw-

ful content,8 and as a result, people who emphasize the legitimate use

of the Internet and others who represent content creators would like to

include anti- piracy or anti- indecency paragraphs in the network neutrality

solution.

12.3 SOME BASICS CONCERNING NETWORK NEUTRALITY

The Internet is characterized by several distinctive features that require

special attention when proposing network neutrality policies. This section

will briefl y outline them.

12.3.1 Congestion on the Internet

When a capacity of a shared portion of the network cannot meet the

demand, it results in congestion and hence yields negative externality.9 In

this case, the market mechanism is unable to produce a socially optimal

resource allocation. In order to improve the situation, some policy inter-

vention can be justifi ed from the viewpoint of social welfare maximization,

provided the social benefi t exceeds the cost.

The demand for network usage is increasing rapidly in the US (Swanson

and Gilder, 2008) and, in fact, in the world at large (Cisco Systems, 2008a,

2008b, 2008c). In the case of Japan, the MIC (2008a) estimates that the

overall traffi c increased by 150 percent from November 2004 through

November 2007 and reached more than 800 Gbps (monthly average,

estimated fi gure) in November 2007; in fact the MIC suggests evidence of

congestion over the trunk network (2007b).


Possible strategies for congestion management can be categorized into

two groups: demand control and supply expansion. Demand control

aims to restrict the demand within the capacity. Possible tools for doing

so include a Pigovian tax, a ‘congestion right’ trading system, and quasi-

property rights granted for certain capacity through auction. Supply

expansion can involve direct investments by the government and fi nancial

incentives off ered to private sectors. Such an expansion can materialize

not only by installing more fi bers, but also by introducing new technolo-

gies, including network cache systems (for example, Akamai) and a new

protocol such as P2P or P4P.10

A problem may arise if a congestion management policy is poorly

designed or used for anticompetitive purposes. If an Internet service

provider (ISP) holds a dominant market position and employs such an

ill- designed congestion management policy, the market cannot independ-

ently remedy the resulting ineffi ciency.

12.3.2 Effi ciency in a Vertically Related Market

Since Internet access providers need content and applications to satisfy

subscriber needs, it is not a priori apparent whether or not such fi rms

utilize their market power in an anticompetitive manner. According to

Farrell and Weiser’s (2003) analysis, the monopolist does not have an

incentive to jeopardize its complementary market at all times. Under the

concept of ‘internaliz[ing] complementary effi ciencies’ (ICE), they point

out that:

in choosing how to license interface information, certify complementors, and otherwise deal with developers, such a fi rm has a clear incentive to choose the pattern that will best provide it or its customers with applications. That is, a fi rm will internalize complementary effi ciencies arising from applications created by others. (p. 101)

They also state that: ‘if a platform monopolist integrates into an adjacent

market, it will still welcome value- added innovations by independent

fi rms’ (p. 102). Similar concerns are shared by Baumol et al. (2007).

It is true that these arguments have clear limitations (Farrell and Weiser,

2003; van Schewick, 2007). For example if the complementary market is

monopolized, the basic ICE framework cannot be applied in the fi rst place

since it assumes that a dominant power resides only on the bottleneck

monopolist side. Moreover, as Farrell and Weiser (2003) claim, when

the situation in the complementary market allows a network monopolist

to execute price discrimination, ICE cannot produce socially effi cient

outcomes. Furthermore the original ICE concept does not consider the


two- sided market structure. In the case where the downstream market

faces Cournot competition, Economides (1998) shows that an upstream

monopoly will ultimately decrease social welfare. Bakos and Katsamakas

(2008) show that the integration of a monopolistic two- sided platform by

one set of users on either side can serve to attain social optimum without

policy interventions.

If the tendency to behave anticompetitively is confi rmed, the possible

network neutrality proposals must include an external mechanism that

disciplines market players.

12.3.3 Quality as ‘the Collectively Produced Commons’

A packet- switching system, in which a certain capacity of the Internet

is ‘shared’ by every user, generates non- rivalry (until it becomes con-

gested) and the non- excludability nature of the communication quality.

Maintaining communication effi ciency under the surge of traffi c should

be considered as constituting ‘the commons’, in which case an individual

network user’s decision- making could cause negative externality. Since

this is a rather classic problem that was already pointed out by MacKie-

Mason and Varian (1998), economic theory already has some solutions

prepared in order to fi nance the optimal size of these goods, including

the classical Lindahl mechanism as well as a recently proposed Falkinger

(1996) mechanism.

Nevertheless, what complicates this simple and rather straightforward

situation is the segmented ownership of the Internet. The Internet is a set

of individual networks, and there exists no controlling authority or fund

allocation system. In order to guarantee a certain level of communication

quality on an end- to- end (E2E) basis, cooperation and cost clearance

among ISPs is essential, which is also mentioned in FTC (2007). There

are three kinds of contracting frameworks between network operators at

present: ‘interconnection’, ‘peering’ and ‘transit’. Among these, intercon-

nection, which was originally applied between facility- based telephone

service providers, has a bidirectional cost clearance system, while the other

two do not. ISPs are connected only through peering or transit. A peering

partner or an upstream ISP in the transit agreement need not bear the cost

of usage accruing at the connecting network. This violates the fairness

of cost sharing or the ‘he who benefi ts ought to pay’ principle unless the

amount of packets is balanced in both directions, which is essential for an

optimal fund distribution. To make matters worse, a rapid spread of bit-

intensive application, especially Internet Protocol (IP) video content, will

upset the balance between upstream and downstream ISPs. Therefore, it is

currently very diffi cult to attain optimal fund distribution.


12.3.4 Other Considerations

There are other factors that should also be taken into consideration when

making policy recommendations: providing broadband Internet as a uni-

versal service or improving the availability of broadband Internet, the cost

of regulation and the ensuing market reaction towards such intervention.

In addition some proponents in the US suggest that stipulating network

neutrality rules and promoting innovation at the edge of the network is

a good option for raising the country’s Organisation for Economic Co-

operation and Development (OECD) ranking.11

12.4 POLICY RECOMMENDATIONS

Although it is important to take non- economic aspects into account as

previously mentioned, this section will basically discuss network neutral-

ity solutions from an economic viewpoint for simplicity reasons, by fi rst

stating the issues from the short- term perspective and then from the long-

term one.

The short- term solution to be discussed below implicitly assumes the

dominance of existing network operators, whereas the long- term one

does not. This may create mutually incompatible solutions in each time

framework, and can explain a certain discrepancy in the network neutral-

ity debates. For example a strong confi dence in wireless innovation and

uncertainty regarding the future industrial organization support Yoo’s

(2006) network diversity claim, whereas the pro- network neutrality side

focuses more on the current status quo.

12.4.1 Short- Term Issues

In the short term, under which network capacity is considered to be fi xed

and new entry to the market is impossible, the issue can be phrased as

follows: How to maximize the static effi ciency? or, How to properly disci-

pline incumbent network operators who control bottleneck facilities?’12 In

this case, we assume that some market power exists in the network market,

while the content and/or application market is more competitive. For sim-

plicity, my discussion will focus on a situation where ISPs are completely

integrated into the network facility operators’ business.

There are two major tasks to be dealt with in order to solve this short-

term issue. First, Task 1 is to prevent anticompetitive behavior on the part

of a network operator. Considering the cost of regulation and informa-

tion asymmetry, competition is always the most desirable solution – that


is, government intervention must be kept to a minimum. In this case, the

aforementioned ICE concept serves to determine the optimal policy mix.

Since ICE assumes that a profi t- seeking bottleneck monopolist acts to

maximize its effi ciency, it can be used as a foundation for policy- makers

to build a least- regulated market. However the eff ective range of ICE has

some limitations. Thus, it is important to distinguish the case where we

can rely on ICE from that where we cannot. On the basis of this considera-

tion, I propose a possible framework that regulators can rely on (Figure

12.1). For example this fl owchart shows that when the situation leads us to

‘Outcome 2’, building up a competitive pressure in the ISP market through

a signifi cant market power (SMP) type of regulation is required.13

There may be some drawbacks in the use of the fl owchart. The most

signifi cant one is that implementation requires a great deal of time and

substantial costs. Due to the heterogeneity of the market structure, deci-

sions must be made on a community- by- community basis. Moreover the

speed of technological change requires a frequent review of such decisions.

Furthermore information asymmetry may cause deterioration of the effi -

ciency of the overall mechanism. Thus it is not practicable to prepare a

Does the market condition guarantee the social optimality

of ICE?*

Can the competition in the network market

improve the situation?

Yes

No

No regulation for net neutrality issues

Outcome 3

Incubate competition by SMP regulation

Outcome 2

Yes

No Net neutrality regulation(New common carrier regulation)

Outcome 4

Does the currentcompetition disciplineincumbent network

operators?

No regulation or laissez-faire

Outcome 1

Yes

No

• Local loop unbundling• Vertical separation between ISPs and network

• Nondiscrimination rule• Traditional entry/price regulation for a bottleneck monopolist if necessary

• Traditional entry/price regulation for a bottleneck monopolist if necessary

Note: * The counterexamples listed in Farrell and Weiser (2003) and van Schewick (2007) provide a good basis for implementing this decision box.

Figure 12.1 Flowchart for selecting an optimal policy mix for a short-

term issue


complete set of policies on an ex ante basis. In other words, policy- makers

should initially trust the initiative of market players and the overall market

mechanism, and then start intervening only in the event of a market failure

that cannot be remedied by private initiatives, and if the social benefi t

of a certain government intervention exceeds its cost. In this case, in

order to minimize the risk of or the uncertainty about the future possible

policy intervention, it is important to declare a general policy guideline

beforehand. Since the primary policy target we are now considering is to

prevent anticompetitive behavior of bottleneck monopolists, I believe it is

appropriate for a government to declare a nondiscrimination principle as

a ground rule. Related authorities (an antitrust court or the quasi- judicial

section or ‘alternative dispute resolution’ of telecom regulators) then

monitor the market to see if any breach is happening. Once detected, they

can follow the fl owchart and build an appropriate policy tool.14

The second- biggest drawback is that taking the ex post antitrust mecha-

nism into consideration, the use of the fl owchart might be considered as

‘gilding the lily’. In the testimony before the US Senate Committee on

Commerce, Science, and Transportation (22 April 2008), Lessing pointed

out that relying on an ex post remedy will stifl e the incentive at present

and choke future innovation. If this is the case, the next step that should

be considered is determining the optimal mix of ex ante and ex post

treatments.

The effi ciency of the aforementioned proposal presupposes the effi cient

use of the current network capacity. Task 2, therefore, entails motivating

network operators to utilize their existing resources effi ciently through

‘reasonable’ network management. The majority of network neutrality

debaters on both sides agree on the necessity of such management; they

disagree, however, on what constitutes ‘reasonable management’.

There is good reason to believe that it is very diffi cult to attain eco-

nomically justifi able ‘reasonableness’ under the current market where the

E2E transmission quality has the characteristics of a commons without a

proper cost- sharing mechanism, and the monthly fi xed charge contract for

subscription is very popular. Because of these, ISPs and end- users do not

have the socially optimal level of incentives for effi cient use of network

resources. Thus it is important to motivate ISPs properly to conduct eco-

nomically justifi able reasonable network managements in the short run

(and to invest suffi ciently in the long run) and, at the same time, to intro-

duce a pricing mechanism that charges a higher fee if the end- user requires

better throughput.

Since the short- term issue has been defi ned above from the viewpoint

of maximizing static effi ciency, one of the theoretical answers to this is an

auction- like mechanism, such as the concept of a ‘smart market’ put forth


by MacKie- Mason and Varian (1998). In an ideal situation, such a mecha-

nism will guarantee economic effi ciency. Alternatively if the unit price level

is somehow properly determined (possibly through trial and error), metered

tariff s may be able to generate satisfactory results; however such tariff s incur

high transaction costs, causing a serious negative impact on the broadband

penetration. Kruse (2008) proposes a similar idea using four types of quality

class, and MIT (2005) examines the possibility of modifi ed fl at rates (that is,

‘pricing tiered by peak rate’ and ‘pricing tiered by traffi c volume’). On the

other hand when operators face enough competitive pressure and suffi cient

disclosing requirements, it might be better to take a laissez- faire approach

and leave it to the market instead. It is important to note that all such sug-

gestions implicitly require some clearing house mechanism on the basis of

the ‘sending party’s network pays’ principle (Kruse, 2008, p. 30).

12.4.2 Long- Term Issues

The challenge faced in the long term, under which it would be possible

to alter network capacity, can be stated as follows: How to maximize

dynamic effi ciency? In other words, the challenge lies in determining how

to motivate network operators to deploy the optimal network capacity

while generating positive externality.

Capacity building of the network can be realized by employing various

methods other than installing more fi bers or routers. Operators can expand

capacity ‘virtually’ by adopting new technology (for example, wavelength

division multiplexing, IP multicast technology), better protocols (for

example, P2P, P4P) and/or better network management. Regardless of

the method actually deployed, it is still true that expanding capacity will

require additional investment, and private operators will not be motivated

without the appropriate incentives in place.

Some network neutrality proponents suggest that capacity expansions

should be solely fi nanced on the basis of the subscription model – that

is, through additional monthly subscription revenues such as quality of

service (QoS) surcharge from end- users, and not from content and/or

application providers. Hermalin and Katz (2007) show that such a prac-

tice will degrade the transmission quality off ered by ISPs. In addition, if

the sum of subscription revenues is not suffi cient, then this proposal will be

utility- decreasing and unsustainable in the long term.

In order to answer the above question empirically, I conducted a survey

of Japanese broadband users in 2007, using an e- mail and web- based

system. Japan is clearly one of the frontrunners in the broadband policy

development and has the most advanced and cost- eff ective broadband

access environment around the world. Thus, I believe that analyzing the


practicability of the above- mentioned pro- network neutrality proposal in

Japan will provide certain guidelines for other nations. The survey period

was from 27 to 30 July 2007; 912 valid responses were collected out of

3000 contracted monitors. The median age of the respondents was 37.3

years and the median annual household income was 696.4 million yen. On

average, the respondents had 92.4 months of Internet experience and spent

4928 yen per month on Internet access. A total of 803 respondents used

broadband (344 used fi ber to the home – FTTH; 349 asymmetric digital

subscriber line – ADSL; and 110 cable Internet) and had 50.1 months of

experience of subscription. Two questions were specifi cally focused on

estimating the respondents’ willingness to pay (WTP) for improving the

QoS over the Internet (Box 12.1).

Following Hidano (1999), the WTP fi gures were estimated by adopting

a Weibull survival model. In equation (12.1), the survival function, S (T ) ,

can be interpreted as a reduced- form description of the probability that an

individual’s WTP is at least as high as T :

S(T ) 5 1 2 G(T ) 5 expa2expa ln T 2 m

sbb, (12.1)

Where G(·) = cumulative distribution function for T , and m and s are the

parameters to be estimated.

When incorporating a sample’s demographic features, the following

survival function is applied:

S(T ) 5 expa2expa ln T 2 b rXs

bb, (12.2)

where X is a vector of demographic variables and b represents the param-

eters to be estimated.

Table 12.1 shows the responses for survey questions 31 and 32. Table

12.2 summarizes the results of the parameter estimations based on the

maximum likelihood estimation method, which were calculated by using a

software package called CVM 2002 (version 1.0).15

The estimation for equation (12.1), presented in the second column of

Table 12.2, indicates that under perfect price discrimination, each respond-

ent would be ready to spend between 452.1 yen ($5.00)16 and 1064.6 yen

($11.77) on average as a monthly surcharge to improve the quality of com-

munication. Assuming that this average WTP can be applied to broad-

band users in Japan – which totaled 26.44 million as of March 2007 – the

total WTP was between 143.4 billion yen ($1.59 billion) and 337.8 billion

yen ($3.73 billion) each year. This is equivalent to 6 percent to 14 percent

of the major fi rms that invested in telecom equipment during FY2006.17

However as the sixth column shows, such estimates are signifi cantly


infl uenced by some demographic variables. The third, fourth and fi fth

columns show diff erences in WTP between subgroups.

On the other hand, 22.2 percent of the investments made by Japanese

major telcos in 2005 were towards trunk facilities.18 The traffi c volume

transmitted over the Internet in Japan increased 2.5 times over the three

years following November 2004,19 and is predicted to double every other

year.20 This prediction is based on a mere extrapolation of the past trends;

however it is in line with the Internet traffi c estimates of Cisco Systems

(2008c). If this trend continues, even assuming the 20 percent annual

increase of subscription,21 the amount of possible WTP would run short

BOX 12.1 QUESTIONS FOR WTP ESTIMATION

Q31 Assume your ISP company is currently considering an investment plan for improving its backbone infrastructure to alleviate the congestion problem and make your broad-band experience less stressful (i.e., faster downloading, uninterrupted streaming), which costs several billion yen and lasts several years. In order to fi nance this plan, your company intends to impose a surcharge of X yen per month. What is your most likely response?

A 1. I approve the plan and accept the required surcharge. 2. I approve the plan but reject the required surcharge. 3. I approve the plan but reject any surcharge. 4. I disapprove of the plan.

Q32 (For those who chose Answer 4 in Q31) What is your reason for disapproving of the plan?

A 1. I do not require broadband content. 2. I am satisfi ed with the current quality, and thus there is no

need for further investment. 3. I am not satisfi ed with the current quality, but there is no

need for further investment. 4. I do not understand the question. 5. Other reasons.

Note: X is set as 500, 1000, 1500, 2000, 2500, 3000, 3500, 4000, 4500, 5000, 5500, or 6000 yen.


within only two to three years (Figure 12.2). Thus it is clear that the

subscription- model proposal of network neutrality regulation supporters

is not sustainable in Japan.

If technological developments cannot cover the diff erence,22 ISPs must

seek some alternate business models such as an advertisement model or a

vertical integration model. Either scheme would ultimately entail heavy

or bandwidth- intensive subscribers to bear the cost of traffi c congestion;

therefore both schemes appear socially fair.23 It is also possible to rely on

tax revenues, provided this is justifi ed by the size of the positive externality

of broadband Internet.

Table 12.1 Responses to Q31 and Q32

Responses to Q31

Surcharge N Would approve the plan and . . . Would not

approve the

plan (%)Pay the

required

surcharge

(%)

Not pay

the required

surcharge

(%)

Not pay

any surcharge

(%)

500 87 9.2 17.2 37.9 35.6

1000 85 5.9 23.5 42.4 28.2

1500 65 3.1 24.6 40.0 32.3

2000 77 1.3 18.2 50.6 29.9

2500 75 2.7 24.0 48.0 25.3

3000 75 2.7 21.3 49.3 26.7

3500 77 0.0 11.7 58.4 29.9

4000 74 2.7 16.2 37.8 43.2

4500 69 2.9 27.5 43.5 26.1

5000 74 2.7 21.6 41.9 33.8

5500 75 2.7 14.7 50.7 32.0

6000 79 1.3 15.2 35.4 48.1

N 29 178 407 298

Responses to Q32

Reasons for disapproval N %

I do not require broadband content. 106 35.6

I am satisfi ed with the current quality, and thus there is

no need for further investment.

62 20.8

I am not satisfi ed with the current quality, but there is no

need for further investment.

48 16.1

I do not understand the question. 69 23.2

Other reasons. 13 4.4


Table 12.2 Estimated parameters for equations (12.1) and (12.2)

Model Equation

(12.1)

Equation

(12.1)

Equation

(12.1)

Equation

(12.1)

Equation

(12.2)

Sample ALL Richer Male Male and rich ALL

N 436 146 164 73 294

Log-

likelihood

−102.16 −41.89 −55.39 −28.28 −68.83

A/C 208.32 87.78 114.77 60.56 151.66

Coeffi cients

s 4.124***

(2.926)

3.873*

(1.939)

3.210**

(2.540)

2.525**

(2.227)

3.061**

(2.884)

μ 3.603**

(2.510)

4.304**

(2.352)

5.169**

(4.968)

5.999***

(7.079)

In (household

income)

1.365**

(2.025)

Gender

(M = 0, F

= 1)

−1.569*

(−1.904)

Marital

status

(married

= 1, other

= 0)

−0.803

(−1.418)

Student

(full time

= 1, other

= 0)

−1.782

(−1.299)

FTTH

(user = 1,

non- user

= 0)

0.600

(1.058)

Constant −3.180

(−0.659)

Mean WTP ¥1064.6 ¥1470.4 ¥1381.8 ¥1377.2

Truncated

mean WTP

¥452.1 ¥593.9 ¥729.5 ¥923.7

Median WTP ¥8.1 ¥17.9 ¥54.2 ¥159.7

Notes:1 Numbers in the parenthesis are asymptotic t- values.2 *** is less than 1%, ** is less than 5%, and * is less than 10%.3 Truncated points are 0 and 6000 yen.4 ‘Richer’ sample includes respondents with above- median household income.


When a new business model is adopted, the problem remaining is again

how to allocate the raised resources to the appropriate network operators.

For example if a single communication needs two ‘peering’ interconnec-

tions between three ISPs, the middle ISP cannot obtain any revenue from

the other ISPs who can directly charge end- users. Since balancing accounts

directly between numerous ISPs is very costly (Farnon and Huddle,

1997), it may be better to introduce an external and centrally controlled

resource allocation system, such as a universal service fund- like external

mechanism. In addition, as mentioned in section 12.3.3, it is necessary to

provide each network operator with the appropriate incentives to upgrade

the network capacity because of the common feature of Internet quality,

which results in a more- than- optimal usage and a less- than- optimal invest-

ment. The challenges that will need to be addressed in the future are moni-

toring the necessity of each ISP’s investment and its actual eff orts and then

compensating with a proper external fund injection.

12.5 CONCLUSION

This chapter shows that network neutrality is not a simple concept, but

has at least six dimensions. Therefore before considering a possible solu-

tion to the network neutrality issues, it is important to specify the focus of

0

100

200

300

400

500

1st 2nd 3rd 4th 5thYear

Inde

x

Minimum for maintaining the current QoS

Original investment

Lower estimate

Upper estimate

Figure 12.2 Simulations for subscription- model sustainability


the discussion on this concept. Further, in considering policy recommen-

dations, one must fully understand some of the peculiarities of the Internet

business, which include the characteristic of sharing some portions of the

network, the motivations of a monopolistic network operator in a two-

sided market, public good- like features of communication quality, and the

lack of a clearing house mechanism between ISPs.

One of the main fi ndings of this analysis is that there exists no ‘one size

fi ts all’ solution when dealing with the short- term economic issue. For

example, in some cases, an SMP- type regulation should be introduced,

accompanied by a certain allowance of network management. To guar-

antee that such network management is reasonable, a ‘smart market’

proposed by Mackie- Mason and Varian (1998) could be a good guideline.

Alternatively the market could be left to ascertain whether certain condi-

tions have been satisfi ed. As for the long- term issue of capacity building,

my econometric analysis indicates that the subscription model may not

be sustainable without extraordinary high hopes for future technological

developments; it is necessary, rather, to introduce other funding sources to

fi nance network expansions. For both short- term and long- term cases, a

fund allocation mechanism must be established.

This chapter might generate more questions than answers. Among

them, three important issues are: (1) implementing these proposals in the

real world; (2) determining the accurate boundaries within which ICE can

guarantee overall effi ciency; and (3) improving the accuracy and robust-

ness of WTP estimations in other samples or settings. These constitute the

areas that need to be explored in future works.

NOTES

1. This research was assisted by a grant from the Abe Fellowship Program administered by the Social Science Research Council and the American Council of Learned Societies in cooperation with and with the funds provided by the Japan Foundation Center for Global Partnership. The survey was supported by a grant from the International Communications Foundation.

2. As Clark (2007) and others have already pointed out, the reason why the network neu-trality issue has become the focal point of discussion in the US is that the government ‘ha[s] abandoned the idea of increasing competition through facilities unbundling’ (p. 704).

3. Keynote Address at the Silicon Flatirons Symposium, Powell (2004). 4. FCC (2005). 5. See Jordan (2007) for details on the seven bills submitted to the 2005–06 Congress. The

historical development of this debate is well summarized by Atkinson (2007). 6. In the US, since the trunk segment of the Internet is relatively abundant in capacity

compared to its access, partially due to the dot- com bubble, the locus of this problem seems currently limited within the cable’s access portion. In Japan on the other hand, where the access segment has a much higher capacity than in the US, congestion may be


observed in the trunk segment. Of course, since NTT’s FTTH is designed such that its access line is shared by a maximum of 32 users, there is a possibility of some congestion at the access line, if bit- intensive real- time applications are excessively used.

7. The discussion concerning the two- sided market is well summarized by Rochet and Tirole (2003, 2006), Eisenmann et al. (2006) and Armstrong (2006).

8. Svensson, P. (2008). 9. Excess demand is not the single cause of congestion. See Gupta et al. (2005).10. ‘P4P’ stands for proactive network provider participation for P2P or provider portal for

P2P. See Xie et al. (2008) for details.11. OECD (2007).12. These are the primary focuses of the MIC’s discussion (2007b); the ideas presented in

the report are fundamentally similar to the policy recommendations discussed below.13. This particular outcome coincides with the situation that Jordan outlines in his article

(2007).14. Bauer (2007) proposes a similar ‘wait and see’ approach.15. A software of the Research Institute for Community Policy Co., Ltd.16. The currency conversion refl ects the exchange rate in January 2009.17. MIC (2007c).18. This fi gure comes from investments by ‘registered telcos’ in trunk line facilities, switch-

ers, routers and computers, coupled with investments by large ‘notifi ed telcos’ in switch-ers, routers, multiplexers, line concentrators and computers (MIC, 2006).

19. MIC (2008b).20. The presentation by Mori (2008) the then Vice- Minister of the Ministry for Internal

Aff airs and Communications (MIC) at 30th Anniversary conference of PTC ’08.21. This assumption is very optimistic. According to the MIC (2007a), Japanese broadband

subscribers increased by 13 percent in 2006 and their growth rate has been decreasing.22. A Japanese ISP, SoftBank, claims that the backbone cost per capacity has been decreas-

ing at the rate of 20–30 percent per annum. (SoftBank, 2008). Assuming that this tech-nological development continues at 20 percent annually, the broadband subscription increases at a more plausible rate of 10 percent annually, and the retail market condi-tion remains constant, we would still need new revenue sources in the 7th to 10th year from the starting point of the calculation.

23. It is possible that the P2P or P4P protocol may create ‘unconsciously’ heavy users and raise some fairness concerns.

REFERENCES

Armstrong, M. (2006). Competition in two- sided markets. RAND Journal of Economics, 37(3), 668–91.

Atkinson, R.C. (2007). Network neutrality overview. Presentation mate-rial from the 9th Annual Telecom, Cable and Wireless Conference, Austin, TX. Retrieved 14 August 2008 from https://www4.gsb.columbia.edu/null/download?&exclusive=fi lemgr.download&fi le_id=646279.

Bakos, Y. and E. Katsamakas (2008). Design and ownership of two- sided net-works: implications for Internet platforms. Journal of Management Information Systems, 25(2), 171–202.

Bauer, J.M. (2007). Dynamic eff ects of network neutrality. International Journal of Communication, 1, 531–47.

Baumol, W.J., M. Cave, P. Cramton, R. Hahn, T.W. Hazlett, P.L. Joskow, A.E. Kahn, J.W. Mayo, P.A. Messerlin, B.M. Owen, R.S. Pindyck, V.L. Smith, S. Wallsten, L. Waverman, L.J. White and S. Savage (2007). Economists’ statement


on network neutrality policy. Retrieved 14 August 2008 from http://www3.brookings.edu/views/papers/litan/200703jointcenter.pdf.

Cisco Systems, Inc. (2008a). The Exabyte era. Retrieved 1 April 2008 from http://www.cisco.com/en/US/solutions/collateral/ns341/ns525/ns537/net_implementa-tion_white_paper0900aecd806a81a7.pdf.

Cisco Systems, Inc. (2008b). Approaching the zettabyte era. Retrieved 11 December 2008 from http://www.cisco.com/en/US/solutions/collateral/ns341/ns525/ns537/ns705/ns827/white_paper_c11- 481374.pdf.

Cisco Systems, Inc. (2008c). Cisco visual networking index – forecast and meth-odology, 2007–2012. Retrieved 11 December 2008 from http://www.cisco.com/en/US/solutions/collateral/ns341/ns525/ns537/ns705/ns827/white_paper_c11- 481360.pdf.

Clark, D.D. (2007). Network neutrality: words of power and 800- pound gorillas. International Journal of Communication, 1, 701–8.

Economides, N. (1998). The incentive for non- price discrimination by an input monopolist. International Journal of Industrial Organization, 16, 271–84.

Eisenmann, T., G. Parker and M.W. Van Alstyne (2006). Strategies for two- sided markets. Harvard Business Review, October, 92–101.

Falkinger, J. (1996). Effi cient private provision of public goods by rewarding deviations from average. Journal of Public Economics, 62(3), 413–22.

Farnon, M. and S. Huddle (1997). Settlement systems for the Internet. In B. Kahin and J.H. Keller (eds), Coordinating the Internet (pp. 377–403). Cambridge, MA: MIT Press.

Farrell, J. and P.J. Weiser (2003). Modularity, vertical integration, and open access policies: towards a convergence of antitrust and regulation in the Internet age. Harvard Journal of Law and Technology, 17(1), 85–134.

Federal Trade Commission (FTC) (2005). Appropriate framework for broadband access to the Internet over wireline facilities. 20 FCC Rcd 14986 (policy state-ment). Retrieved 4 January 2009 from http://hraunfoss.fcc.gov/edocs_public/attachmatch/FCC- 05- 151A1.pdf.

Federal Trade Commission (FTC) (2007). Broadband connectivity competition policy: RTC staff report. Retrieved 14 August 2008 from http://www.ftc.gov/reports/broadband/v070000report.pdf.

Gupta, A., D.O. Stahl and A.B. Whinston (2005). Pricing traffi c on interconnected networks: issues, approaches, and solutions. In S.K. Majumdar, I. Vogelsang and M.E. Cave (eds), Handbook of Telecommunications Economics (Vol. 2, pp. 413–39). Amsterdam: Elsevier.

Hermalin, B.E. and M.L. Katz (2007). The economics of product- line restrictions with an application to the network neutrality debate. Information Economics and Policy, 19(2), 215–48.

Hidano, N. (ed.) (1999). Kankyo to gyousei no keizai hyouka: CVM manyuaru [Economic evaluation of environments and policy: manual for CVM analysis]. Tokyo: Keiso Shobo.

Jordan, S. (2007). A layered network approach to net neutrality. International Journal of Communication, 1, 427–60.

Kruse, J. (2008). Network neutrality and quality of service. Intereconomics, 43(1), 25–30.

MacKie- Mason, J.K. and H.R. Varian (1998). Economic FAQs about the Internet. In L.W. McKnight and J.P. Bailey (eds), Internet Economics (pp. 27–62). Cambridge, MA: MIT Press.


Ministry of Internal Aff airs and Communications (MIC) (2006). Survey on communications industry, August 2005: investment March 2006. 17 February. Retrieved 6 January 2009 from http://www.johotsusintokei.soumu.go.jp/statis-tics/pdf/HS200513_001.pdf.

Ministry of Internal Aff airs and Communications (MIC) (2007a). Information and communications in Japan. White Paper 2007. Retrieved August 2008 from http://www.johotsusintokei.soumu.go.jp/whitepaper/eng/WP2007/2007- index.html.

Ministry of Internal Aff airs and Communications (MIC) (2007b). Network no churitsusei ni kansuru kondankai houkokusho [Report on network neutral-ity]. Retrieved 14 August 2008 from http://www.soumu.go.jp/s- news/2007/ 070920_6.html#bt.

Ministry of Internal Aff airs and Communications (MIC) (2007c). Basic survey on the communications industry. 30 November. Retrieved 6 January 2009 from http://www.johotsusintokei.soumu.go.jp/statistics/pdf/HB200600_001.pdf.

Ministry of Internal Aff airs and Communications (MIC) (2008a). Wagakuni no Intaanetto ni okeru torahikku no shukei keisan [Estimation of traffi c volume on the Internet in Japan]. Retrieved 14 August 2008 from http://www.soumu.go.jp/s- news/2008/080221_3.html.

Ministry of Internal Aff airs and Communications (MIC) (2008b). Tabulation and estimation of Internet traffi c in Japan: announcement of tabulation results as of November 2007. 21 February. Retrieved 14 August 2008 from http://www.soumu.go.jp/s- news/2008/082221_3.html.

MIT (2005). The broadband incentive problem. A white paper prepared by the Broadband Working Group MIT Communications Futures Program (CFP) and the Cambridge University Communications Research Network. Retrieved 14 August 2008 from http://cfp.mit.edu/docs/incentive- wp- sept2005.pdf.

Mori, K. (2008). ICT policy in Japan. Presentation material at the 30th Anniversary conference of PTC ’08, Honolulu, Hawaii, 13 January.

Organisation for Economic Co-operation and Development (OECD) (2007). OECD broadband statistics to June 2007. Retrieved 6 January 2009 from http://www.oecd.org/sti/ict/broadband.

Peha, J.M. (2007). The benefi ts and risks of mandating network neutrality, and the quest for a balanced policy. International Journal of Communication, 1, 644–68.

Powell, M.K. (2004). Preserving Internet freedom: guiding principles for the indus-try. Keynote address at the Silicon Flatirons Symposium. Retrieved 4 January 2009 from http://hraunfoss.fcc.gov/edocs_public/attachmatch/ DOC- 243556A1.pdf.

Rochet, J.C. and J. Tirole (2003). Platform competition in two- sided markets. Journal of European Economic Association, 1(4), 990–1029.

Rochet, J.C. and J. Tirole (2006). Two- sided markets: a progress report. RAND Journal of Economics, 37(3), 645–67.

SoftBank (2008). Dai yon kai intaanetto seisaku kondankai purezenteishon shiryo [Presentation material for the 4th meeting of panel on Internet policy]. 27 May. Retrieved 26 December 2008 from http://www.soumu.go.jp/joho_tsusin/poli-cyreports/chousa/internet_policy/pdf/080527_2_si4- 4.pdf.

Svensson, P. (2008). Peer- to- peer networks go legit. 14 March. Retrieved 23 December 2008 from http://www.msnbc.msn.com/id/23636873/.

Swanson, B. (2007). The coming exafl ood. Wall Street Journal, 20 January. Retrieved 18 April 2009 from http://online.wsj.com/article/ SB116925820512582318.html.


Swanson, B. and G. Gilder (2008). Estimating the exafl ood: the impact of video and rich media on the Internet – a ‘zettabyte’ by 2015? Discovery Institute. Retrieved 14 August 2008 from http://www.discovery.org/a/4428.

van Schewick, I.B. (2007). Towards an economic framework for network neutral-ity regulation. Journal of Telecommunications and High Technology Law, 5, 329–91.

Wu, T. (2003). Network neutrality, broadband discrimination. Journal on Telecommunications and High Technology Law, 2, 141–75.

Xie, H., A. Krishnamurthy, A. Silberschatz and Y.R. Yang (2008). P4P: explicit communications for cooperative control between P2P and network providers. Retrieved 14 August 2008 from http://www.dcia.info/documents/P4P_Overview.pdf.

Yoo, C.S. (2006). Network neutrality and the economics of congestion. Georgetown Law Journal, 94, 1847–1908.

PART V

Interdependent innovations and regulatory policies: mobile network deployment and mobile Internet developments

275

13. Interdependent innovation in telecommunications: risk, standardization and regulation1

Bruno Basalisco, Andy Reid and Paul Richards

13.1 INTRODUCTION

This chapter is concerned with developments within and between three

distinct strands of the telecoms industry, namely those of standardiza-

tion, the development of technology policies and the regulation of telecom

operators who are attributed some level of market power and subject to

sectoral regulatory rules. It is taken as a self- evident fact that increasingly

in the information and communication technology (ICT) sector, the com-

mercialization of innovatory technologies which is the underlying aim of

all parties to harness the benefi ts of technical progress, requires the coor-

dination of activities across many parties and even diff erent sectors. This

chapter explores the issues arising for what might be described as good

regulatory practice given the context of developments in standardization

bodies and technology policies.

Regulation and innovation are not easy bedfellows. In fact, technologi-

cal or business change adds much complexity to the dialogue between the

state and the market which is the essence of regulation. The latter can have

an indirect eff ect on fi rms’ incentives to innovate via its impact on the level

of competition in an industry.2 Moreover, regulatory policy can directly

aff ect the path of technological development – and be shaped by it.3

Because of the converging nature of technological change in the ICT

industries, the interplay of regulation and innovation trespasses the

boundaries of industries subject to regulatory oversight. Regulators,

especially those showing an interest in more effi cient forms of regulation

(also known as re- regulation), are increasingly aware of models of innova-

tion in network industries (Farrell, 2003) and are also keen to understand

the functioning of highly innovative unregulated industries. On the other

hand, innovations in technologies and businesses across the ICT industries


have long been shifting the boundary between competition and regula-

tion, which infl uences the whole set of public policies aimed at redressing

market failures (Farrell and Weiser, 2003).

Industry players and public policy alike benefi t from the highest possi-

ble degree of regulatory certainty. This implies that regulatory practice has

to operationalize the innovation dynamics and its impact on competition

in regulated markets. A key procedure in telecommunications regulation

is the defi nition of markets, which underlies any analysis of market power

and the potential remedies to address the latter.4

The concept of emerging markets has been introduced by regulators

in order to account for the interplay of innovation and regulation in the

procedure for market defi nition. It is a subject of intense debate among

academics and practitioners (Dobbs and Richards, 2004; Crocioni, 2008).

Innovations in technologies and business foster new products and serv-

ices, which can provide a challenge for quantitative market defi nitions

(informed by demand substitutability, which suff ers in this case from high

levels of uncertainty). In fact, regulators often acknowledge that the devel-

opment of these innovations could be artifi cially shaped by regulation. As

a result, regulators at times opt to forebear from these innovative products

and services, labeled as part of emerging markets. We believe that the issue

of emerging markets is of high relevance because it lies at the interface of

two fi elds: the specifi c context of applying ex ante market power remedies

in telecommunications regulatory regimes, but also the bigger picture

of the technology policy discourse on the promotion of investment and

innovation.

Specifi cally, the chapter discusses how regulatory policy- making could

stand to benefi t from a framework for emerging markets which accounts

for the extent to which each innovation depends on the strategic interplay

of many parties such as network operators, equipment manufacturers,

regulators and policy- makers. Our analysis stresses that current regulatory

theories and practice underappreciate the importance of the interdepend-

ent nature of innovation (in its risks and benefi ts). We perform a brief

analysis of a set of new products and services spurred by communications

network innovations and discuss the relative nature of innovation. In light

of this and in a spirit of constructive criticism, we argue that there is con-

siderable merit in operating the regulatory framework in such a way that

it signals that diff erent scenarios based on the nature of innovation will

indeed be treated diff erently in regulatory terms guided by the degree to

which the innovation is risk- bearing within and between diff erent parties.

The chapter is structured as follows. Section 13.2 below sets the scene of

the nature of interdependency of innovation for individual fi rms. Section

13.3 discusses the merits and shortcomings of two existing responses to the

Interdependent innovation in telecommunications 277

challenge of interdependent innovation in the communications industries:

standardization and technology neutrality. Section 13.4 reviews devel-

opments of some examples of innovation of network technologies and

their implications on communications services. Section 13.5 concludes by

drawing some implications of our perspective on regulatory practice and

specifi cally in the context of market reviews and the assessment of market

power.

13.2 THE INTERDEPENDENT NATURE OF INNOVATION

Many fi rms need to vault over company boundaries in order to improve

their products and services – a phenomenon popularized as ‘co- opetition’

(Brandenburger and Nalebuff , 1996). Even large, dominant industry

players have realized those size and market shares are not suffi cient to be

able to leverage innovation processes in their industry. Some fi rms eff ec-

tively manage innovation platforms. This is for instance the case of Intel’s

eff orts in orchestrating innovation across its industrial ecosystem through

carefully crafted policies of entry and intellectual property rights (IPR)

sharing in complementary markets (Gawer and Henderson, 2007).

When considering the role of interdependence in aff ecting innovation

eff orts, an important aspect is whether systems present modularity in their

components (Baldwin and Clark, 2000; Langlois, 2002; Farrell and Weiser,

2003). Many fi rms incorporate technological change in their competitive

strategies. Henderson and Clark (1990) highlight that some fi rms exploit

their knowledge of the very architecture of systems by leveraging linkages

between core concepts and components. They thus suggest taking a more

nuanced perspective of the traditional incremental versus radical innova-

tion dichotomy by accounting for architectural and modular innovation.

The competitive advantages of a fi rm may depend on innovation proc-

esses involving its co- opetitors – defi ned as their network of suppliers,

users and complementors (Afuah, 2000). These can be a source of innova-

tion, lead users themselves of the technology, and providers of information

useful to refi ne innovation. Moreover, many fi rms engage in collaborative

ventures with other industry actors with the intent of coordinating the

development of innovation – such as industry fora, joint ventures and col-

laborative agreements. As the rate and uncertainty of technological change

vary across sectors, diff erent industrial ecosystems are denoted by distinct

forms of co- opetition. Each of these presents elements which show how

fi rms’ relationships may have adapted to the extent of innovation interde-

pendence to deliver maximum benefi t to their participants (Robertson and


Langlois, 1995). Telecommunications fi rms setting improved standards

via the International Telecommunication Union (ITU), and Intel promot-

ing better designs in the computer industry, are both instances of invest-

ment strategies which minimize the overall absolute level of risks of the

entire production chain and not just of one part.

In fact, in industries such as telecommunications, network, equipment

and protocol design infl uence the degree of standardization and intercon-

nectivity. Further to that, regulatory decisions can potentially contribute

to freeze the industry ecosystem and business models. On the one hand,

the presence of path- dependence implies that successful innovation by one

or more player(s) is likely to infl uence future technological choice across

the industry. On the other, the existence of direct and indirect network

eff ects provides a great economic incentive for diff erent networks to inter-

connect. All of the above add considerable complexity to the exploitation

of network- based innovations.

While international agreements between telecommunications fi rms

serving diff erent markets have been commonplace, on the other hand, the

relationships between players in regulated industries have traditionally

been adversarial, as new entrants accessed key incumbents’ assets through

regulatory supervised processes. Nonetheless, technology- driven changes

in industry architecture and regulatory reform have fostered an increase

in co- opetition.5 Cooperation at a highly granular level between hitherto

rivals is therefore increasingly acknowledged as necessary for industry-

wide innovation and even mandated as part of sectoral regulation. Thus,

the changing role of standardization will be analyzed in greater detail in

section 13.3.1.

Fransman (2007) analyzes ICT ecosystems across some of the world’s

leading regions (US, Japan, Europe). He argues that the dominant para-

digm in telecommunications regulation conceives the process of innova-

tion as an endogenous but separate process from the competitive dynamics

analyzed for the purpose of regulatory decision- making. In general, while

the consequences of innovation are frequently taken into account by

regulators, a failure to acknowledge how that came about in the fi rst place

leaves a key gap in the analysis.

Specifi cally, the telecommunications industry presents several challenges

to welfare- maximizing policy- makers. Static and dynamic economies of

scale and scope favor oligopolistic structures over atomistic competition.6

Moreover, large players’ investment decisions can have spillover eff ects

towards future choices (as frequent in network industries), leading to

path- dependence. Furthermore, this is compounded by technology inter-

dependence – as discussed throughout this chapter.

In practice, interoperability and complementarity can be pivotal


in allowing a technology to succeed. To take a recent example,

Gunasekaran and Harmantzis (2008) show that the commercial viabil-

ity of Wi- Fi could be materially enhanced by the development of what

might have been regarded as competing technologies such as WiMAX

(Worldwide Interoperability for Microwave Access). Frequently it is

not apparent even at the stage of technical development and even of

preliminary deployment, which eff ect will dominate – complementarity

or substitution – and whether the descending services will be a net sub-

stitute or expander of market demand (Richards, 2007). Thus, regula-

tory decisions (on spectrum, access, and so on) can aff ect the path of

technological development. It is evident that regulators are well aware of

these complexities and some of them have adopted an approach which

is intended to be ‘technology- neutral’. Section 13.3.2 presents how this

policy principle was introduced in European Union (EU) law and dis-

cusses its implications.

13.3 BUSINESS AND REGULATORY RESPONSES TO THE INTERDEPENDENT NATURE OF INNOVATION IN COMMUNICATIONS TECHNOLOGIES

13.3.1 The Importance of Risk Mitigation for Operators and Regulators

A key feature of the interdependent risk faced by innovators is the lack

of information about how other players will act in response to a specifi c

innovation. Where there is a strong mutual interdependency, this lack of

information can escalate the amount of risk involved. This is especially

true in telecommunications where interconnection is at the heart of the

industry.

The key challenge for telecommunications fi rms attempting to innovate

is how to assess project- specifi c risk, considering that information about

future returns is unknown at the time of the decision to invest. This is

essentially the defi nition used in the real options approach to the analysis

of risk.7 As a consequence, it is natural for players in the telecommunica-

tions industry, if possible, to defer investment until the reaction of other

players is better known.

A discussion of risk mitigation necessarily falls into the more general

debate on the nature of uncertainty and risk, informed by the seminal

work of Knight (1921). While a comprehensive discussion of risk and

its theories is beyond the scope of this chapter, our study aims (with an

eye to the context of regulatory practice) to highlight a dimension of the


innovation process – its interdependence – which informs the nature of

the uncertainty and the extent of risk associated to it. Because we take a

fundamentally qualitative approach, in what follows we will use the term

‘risk’ in its broadest meaning and this is also in line with the defi nition

of risk which is currently used in regulatory parlance (for example, see

Ofcom, 2005a).

This section discusses the eff ectiveness of standards in reducing interde-

pendency risk between players and the attempt by regulators to set policy

which is independent of specifi c technical innovations. While the former

is a long- established approach, the latter is a relatively recent phenom-

enon.8 First, section 13.3.2 analyzes how, throughout the long history

of telecommunications, technical standardization has been a mechanism

aimed at mitigating the investment coordination problem and highlights

how shifts in the industry architecture are curtailing the eff ectiveness of

this approach. Second, with the introduction of regulated competition

into the industry in the 1980s and 1990s, the reaction of the regulators to

innovation has been another unknown for players deciding on innovation

investment. In response, EU regulators have therefore attempted to make

regulatory policy stand above specifi c technology innovation by being

technology neutral, which is discussed in section 13.3.3.

13.3.2 Response by Firms: the Case for Standardization

Many sectors of the economy present innovation which is associated with

some elements of standardization. Even though this is the case, often prod-

ucts or services do not directly impinge on each other to be commercially

or functionally viable. For example, car manufacturers may need to meet

emission standards and conditions on technical performance but each car

can achieve these goals to a large extent as a stand- alone system.9 The tele-

communications industry is characterized by features which mark it apart

from many other sectors, even other utility sectors such as energy: the

strength of network eff ects and their global scale create a huge economic

incentive for interconnection between the telecommunications networks

of diff erent organizations. However, as we have shown in the previous sec-

tions, this introduces a considerable extra complexity in the exploitation of

an innovation based on the application of new technology.

Firms attempt to mitigate mutual risks through the process of stand-

ardization. Historically, fora such as the ITU have played a strong part.

This allows many parties to precommit to technical standards which will

lessen the risks of failure for networks to interconnect. Transparency and

commitment to standards also means that purchasers of services are less

likely to be reluctant to contract with suppliers.


Industry evolution created a further motivation for agreeing technical

standards. There is now a widespread split between equipment manufac-

turing and network service provision. Technical standards therefore serve

an additional purpose in negotiating this vertical split in the production

chain, allowing a reasonably open market between the equipment suppli-

ers and network service providers. This is often termed interoperability

between equipments as oppose to interconnectivity between network

operators.10

The case for standards would seem to be very strong with advantages

to both network operators and equipment manufacturers. However,

the actual picture is more complex. The economic origins of equipment

interoperability are less clear than the origins of network interconnectiv-

ity. On the one hand, the split was the product of regulatory and antitrust

action in the late 1980s in both the US and in Europe. On the other hand,

the broadening scope of the industry was also driven by the technologi-

cal developments taking place. For example, both Internet Protocol (IP)

and Ethernet technologies, on which more of today’s telecommunications

network depend, emerged from standards bodies which were not at the

time associated with the telecommunications industry.

The involvement of many new parties in the value chains associated

with convergence in the telecommunications sector has complicated the

standards- setting process. There is a proliferation of bodies producing

‘standard’ specifi cations which are frequently overlapping. In extreme

cases, some specifi cations are inconsistent or incompatible between diff er-

ent bodies. Thus some industry fora today focus most of their eff orts not

on writing standards per se, but on profi ling standards created by other

bodies into interoperability agreements. Strategic forum shopping by

industry players complicates matters.

Network operators, in a competitive marketplace, are interested in

innovation as a means of competitive advantage. However, as no single

service provider (SP) can aspire to own and operate a complete global

infrastructure- based network, interconnection is essential. This means

that the exclusive exploitation of innovation is often diffi cult. For network

operators, the details of a standard are often not important diff erentia-

tors between their off ers, and they normally have a strong mutual interest

in a single standard in order to achieve interconnection and competitive

supply of equipment.11 Today, most network operators regard techni-

cal standards as something they ‘consume’ as part of buying equipment.

Several operators have therefore withdrawn from the standards- setting

process with an eye on short- term cost savings. This exacerbates con-

fl icts between equipment manufacturers, where instead operators could

play a welfare- enhancing role by promoting unique standards. Instead,


frequently standards with ‘options’ are selected as a means of accommo-

dating irreconcilable commercial interests.

On the other hand, for the equipment manufacturers, the technical

details of the standard are likely to aff ect their time to bring innovations

to market as well as the market perception of whether they are a ‘technical

leader’ or a ‘technical follower’. Equipment manufacturers are certainly

interested in innovation as a mean of competitive advantage. For example,

in the extreme, all network operators could achieve full interconnectiv-

ity by all buying the same equipment from the same manufacturer. In

this case, manufacturers would be able to exploit fully the value of their

innovations since there would be no need for interworking with other

equipment manufacturers. However, operators share a concern to main-

tain a level of competitive supply and tend to be cautious about exploiting

an innovation which locks them into a supply from a single equipment

manufacturer. This means that the equipment manufacturers can have

ambiguous motivations for open standards. In its turn, these confl icting

motivations may, at least in part, explain the fragmentation in standards

and the multiplicity of standards bodies.

Regulators are interested in standards as a prerequisite for regulated

interconnection. However, even more than network operators, they cur-

rently tend to be concerned about the simple presence of standards rather

than the technical details and, as a result, many regulators tend not to

attempt to follow the detailed technical developments in standards bodies.

As an additional note, the unprecedented rate of architectural change

in telecommunications networks also poses a strong challenge to stand-

ardization eff orts. In the past, technological innovation has been confi n-

able to a reasonably static overall architectural framework. For instance,

when digital technologies made electromechanical exchanges and ana-

logue transmission systems obsolete, the overall architectural structure

of ‘access’, ‘switching’ and ‘transmission’ was unaff ected. Also, the basic

services running across the network were not aff ected by this innovation,

thus leaving the pricing and the industry business model unchanged.12

Conversely, in the present environment, the nature of basic services, the

pricing of these services – as well as the industry business model – are all

changing under the pressure of technological developments. It is conse-

quently hard for the standards bodies themselves to keep their own organi-

zations and membership in step with the pace of the architectural change

in the broader industry.

To conclude, we have shown that the real picture of standards is more

complex and more ambiguous than a simple investment coordination

device. The standards- setting process is certainly alive and well; however,

its eff ectiveness is more nuanced.


13.3.3 Response by Policy- Makers: Technology- Neutral Regulation in

the EU

The principle of technology- neutral regulation refl ects a trend in the

policy- makers’ stance on the role of regulation in the face of fast and

hard- to- predict technological change. The concept started being discussed

in EU circles around a decade ago. Currently, it is a high- level princi-

ple enshrined in the 2002 Electronic Communication Network Services

(ECNS) framework.13 It has been instrumental to the achievement of the

objective to strengthen the EU internal market, by aiming to eliminate

some barriers that, in the face of convergence, prevented a comprehen-

sively similar treatment of communications operators throughout the

EU Member States. Nonetheless, technology neutrality is continuously

traded off with other high- level principles associated to both the ECNS

framework and other EU policies (for example, the i2010 industrial policy

initiative in European Commission, 2005). While this principle has wide-

spread recognition industry- wide,14 countries like Korea have shown that

a policy choice of a diff erent trade- off between regulatory, competition

and industrial policy can be fruitful.15

Policy- makers, understanding of this principle has kept evolving and

gaining nuances while communications industry players gained a more

sophisticated understanding of its technology dynamics, taking on board

the history lesson that technological progress can take unexpected paths.

This disrupts fi rms’ large- scale plans as well as aff ecting even the best-

meaning regulatory eff orts. For instance, integrated services digital network

(ISDN) technology, much hyped in the late 1980s–1990s, was expected to

be a neutral technology since its network would have been multipurpose

and carried a variety of services. This would have translated into a reduced

need for regulation (see the 1987 EC telecommunications Green Paper, pp.

21, 33).16 Ten years on, it had become more evident that a single techno-

logical development (in the form of ISDN) could not live up to its promise

to radically reshape consumers’ demand and regulation. This contributed

to awareness amongst policy- makers that a technology- neutral regulatory

framework was needed, rather than just a neutral technology (see the 1997

EC convergence Green Paper, pp. 19–21). This resulted above all in the

harmonization eff orts in the post- 1998 framework review.

We see technologically neutral regulation as having two distinct funda-

mental connotations. On the one hand, technology neutrality is a feature

of the regulatory toolbox. It allows regulatory tools to adapt to a con-

verging environment and be applicable for general regulatory purposes.

It refl ects regulators’ aim to build and deploy the analytical tools needed

for future- proof regulation, given that technology and markets present


underlying risks. This is the micro side of the principle. On the other hand,

taking a macro view of the economy, the principle is also leveraged to

justify a pilatesque stance on regulatory- mediated technological choice.

This is the case the more regulatory decisions enter the fi eld of industrial

policy. The principle can thus be seen as a facet of a non- interventionist

industrial policy.

This puzzle refl ects the underlying conundrum regulators face between

static and dynamic effi ciency, which at times need to be traded off , and

this tension is also evident in the ECNS framework.17 The principle

of technology- neutral regulation refl ects regulators’ desire to achieve

both static (level of investments) and dynamic (innovation) effi ciency.

Technology- neutral regulation is relatively unproblematic where the state

and outlook of markets are unequivocal. It matters most where risk adds

to the complexity and may require less orthodox analytical approaches.18

Regulatory oversight shapes and is shaped by the technological profi le of

industries.19 This is because in regulated industries, investment decisions

are informed by the national regulatory authority (NRA) approach to

risk. Both the NRA’s current orientation and its past legacy of access poli-

cies will be relevant, a form of regulatory path- dependence.

The current EU regulatory framework faces the challenge of overcom-

ing its underappreciation of the relevance of new services and markets,

driven by technological progress.20 Technology neutrality is an amelio-

rative approach to regulation insofar as the acknowledgment of path-

dependence makes regulators assess that technological choices do matter

and their dynamic eff ects yield more complexity to policy decisions. On

the other hand, if in practice the technology neutrality principle leads

regulation to hinder fi rms’ welfare- enhancing attempts to mitigate the

interdependent risks of technological change, then the application of this

principle should be constrained by the greater industrial and technology

development societal imperative.

Across the world, institutions and policy- makers draw several lines

between competition and regulatory policy on one side and technology

and industrial policy on the other in diff erent ways, which impacts the

competitiveness of their economies. Technology- neutral regulatory policy

stands on one of those boundaries. Whether regulators focus on neutral-

ity to technological competition (within the market) or technological

development (for new markets) will depend on what view they will take

of Schumpeterian competition. In practice it will depend as well on what

tools will be available for regulators to analyze and address market failure

in such a context, as will be discussed in the following sections.

In conclusion, the analysis of the origin and rationale of the concept

of technology neutral regulation suggests a series of points. First, the


importance given to technology neutrality as one of the high- level prin-

ciples of the EU communications regulatory framework shows that

policy- makers are keenly aware that technological development in the

communications space is not a one- player game: if it were so, straightfor-

ward regulatory measures could ensure the achievement of policy objec-

tives. Second, because of its open- endedness, the concept of technology

neutrality risks being used to mask the ambiguity related to the confl ict-

ing nature of regulatory related remits. This is because regulators can at

times de facto perform industrial policy functions, although the regulatory

rhetoric can underplay these for political convenience. Third, successful

regulation thrives on a level of independence from government, as well as

to be accountable to its stakeholders (Majone, 1996); thus technology and

industrial policy is an uncomfortable territory for regulators to tread in.

These considerations may lead us to discard technology neutrality as

insubstantial insofar as it cannot consistently be refl ected in regulatory

practice across the board (while specifi c regulatory measures can be tech-

nology neutral, for example, spectrum allocation). Nonetheless, we argue

that a healthy debate on this principle can have a benefi cial eff ect on soci-

etal welfare if it can help reassess the future role of regulation and indus-

trial policy and the division of labor within the public sector in its task to

serve societal interests, which at times are confl icting.

13.4 RISK AND INNOVATION IN NETWORK TECHNOLOGIES

13.4.1 A Simple Model of Risk Bearing in Network Innovations

In section 13.3, we saw that while network operators and equipment

manufacturers undertake standardization eff orts and regulators attempt

policies which are technology neutral, neither are perfect instruments.

Innovation normally involves several aspects of risk. This could arise

from any number of things which may be unknown at the time a decision

is made to invest in an innovation, such as technology risk, deployment

risk and market risk,21 where consumers do not purchase the service in

anticipated volumes. Moreover, a single organization – such as a service

provider (SP) – is seldom able to internalize fully the risks of successfully

developing and deploying a technological innovation, independently from

other parties.

In this section we analyze several historical technological innovations in

telecommunications networks and then apply the analysis to a problematic

recent example of the application of regulation to a telecommunications


service – Ethernet services. Our simple model focuses on the perspective

of a single SP, the business of which is infl uenced by the innovation dis-

cussed. Our model separates and measures innovation risk along two com-

ponents. The fi rst component refl ects the degree of risk occurring inside

the SP domain, therefore under its control (the vertical axis in Figure

13.1). The second component (horizontal axis) measures risks which fall

outside the domain of our reference SP.22 These outside risks arise from

relationships of our reference SP (the innovator) to other organizations

which may be at a diff erent level in the value chain or alternatively direct

competitors at the same point in the value chain. Where this component

is high, our SP requires a great extent of coordination with those external

organizations in order to gain benefi ts from the innovation.

Figure 13.1 represents a set of technological innovations introduced in

the past in the communications industry. Technologies that are positioned

close to one of the axes are those where an SP could develop or deploy

the innovation independently or gain benefi t at arm’s- length from an

externally developed or deployed innovation. In both cases our reference

SP is able to internalize fully all risks associated to the change in network

technology. Instead, the more the innovation is depicted away from the

axes, the higher will the level of mutually dependent risk be. Notice the

three scenarios associated to the nature of risk, which we will later argue

are relevant to regulatory assessments.

Given the telecommunications industry features of network externalities

Risks only to

external orgs

Risks external to service providerHighLow

Risks only to SP

Ris

ks

inte

rnal

to s

ervic

e p

rovid

er

Low

Increasing risk from

mutual interdependency

High

Figure 13.1 Basic model of interdependency risk


and path- dependence, our simple model allows us to take into account

that an SP which has market power may be able to innovate in a way

which is not risky for itself but imposes risks on others. We can then

discuss whether such a fi rm may have a strategic interest to do so, which

carries both business and public policy implications (depending on the

context). In fact, as the following examples show, there is a huge variety

of circumstances in which the eff ects of innovation can assist third parties,

sometimes almost inadvertently.

In the 1980s, the whole telecommunications industry underwent a huge

technological change with the introduction of digital technology (digitali-

zation). However, this did not change the basic public switched telephone

network (PSTN) service used by most end- users, nor was it necessary for

all service providers to change technology in concert. As a result, any one

service provider could develop and deploy digital switching within their

own control, independently of any changes undertaken by their custom-

ers. While this technical innovation carried very large technology risk,

there was little mutual dependence risk. As a result we place this innova-

tion high but near to the y axis.

Conversely, the dramatic rise of the Internet and its applications such as

the World Wide Web (WWW) and e- mail in the 1990s grew as independ-

ent users of telecommunications network outside the domain of the tel-

ecommunications service providers. While the Internet has had a dramatic

impact on end- users who have bought personal computers (PCs) in order

to access it, the telecommunications service providers simply provided

their existing services – leased lines and PSTN calls for modem dial access.

Again there was little mutual dependence risk and we place this innovation

on the right side but close to the x axis.

However, the introduction of mobile telephony in the late 1980s required

a completely new network infrastructure as well as a completely new set

of end- user equipment. The end- user equipment players and the network

operator players had a strong mutual dependency. Digital mobile saw a

major standardization initiative in Europe – Global System for Mobile

Communications, GSM – which was ultimately highly successful for the

players involved. With the high level of mutual dependency, we place this

towards the top right- hand corner.

At about the same time, an even bigger standardization project was

also being undertaken – Broadband ISDN (B- ISDN23). This too required

mutual coordinated innovation by both the network operators and the

end- user equipment industry. This too we place towards the top right-

hand corner. However, this project essentially failed. Instead the end- user

equipment (PCs) chose to follow the Internet standards, not the B- ISDN

standards, a critical reason being that the end user could access Internet


service independently of a network operator’s investment in innovation.

In essence, the availability of an alternative without mutual dependency

risk killed B- ISDN.

Optical transmission developed rapidly in the 1990s and is an interesting

case of ‘outside’ risk, although not to the same extent as for mobile and

B- ISDN. Initially, the development and deployment of the technology

was entirely ‘inside’ a service provider’s domain and so would have sat

fi rmly on the left- hand axis. However, the nature of optical transmission

costs has radically changed the cost structure of providing telecommuni-

cations services. This enabled changes in the pricing of telecoms services,

which in its turn facilitated market expansion and competition in the

provision of services.24 Moreover, the underlying change in cost structures

aff ected optimal network topology design, which reshaped interconnec-

tion costs, thus destabilizing business models and regulatory regimes. Not

all industry players reacted fast to such spillover from technology, due to

a certain degree of business model and regulatory path- dependence. In

many instances, while the cost structures had changed with optical trans-

mission, tariff s and regulation failed to adapt to these changes, leading to

suboptimal investments by third parties.25 In our fi gure, since the optimal

network design for our reference SP depended on demand from intercon-

nected parties (thus on their network design and pricing), the innovation

of optical transmission is represented as distanced from the vertical axis.

Furthermore, another major innovation of the 1990s in telecommunica-

tions was the Intelligent Network (IN). This allows the fl exible creation

of services such as the 800 service, premium rate, and Ring Back When

Free. While there is a certain amount of mutual dependency between the

network operator and its customers for these technology services, they

were developed with little overall risk to either the service provider or the

end- user. It is therefore located close to the axes of origin.

Finally, the recent example of asymmetric digital subscriber line (ADSL)

technology for multimedia broadband in the residential market gives a

particular insight. It shows that the staged exploitation of a technological

innovation can reduce overall risk (see Figure 13.2). In fact, ADSL- based

service requires investment from both the SP and end- users in the technol-

ogy, leading to some mutual dependence. However, ADSL evolved as a

small technology increment from a previous version of the technology,

high bit rate DSL (HDSL). This was developed as a means of reduc-

ing the costs of delivering leased line products and was deployed wholly

‘inside’ the SP network without making any change to those products.

This deployment absorbed the great majority of the technology risk, and

especially of technology development which ADSL was able subsequently

to leverage. The kind of dynamic shown by this technology is not unique


to ADSL and will in fact be part of the development of Ethernet services.

In its turn, we highlight that this dynamic bears implications on regulatory

practice, which, as the Ethernet case below shows, has at times yielded

confl icting outcomes to similar regulatory problems.

13.4.2 The case of Ethernet: the Nature of Innovation and Regulatory

Practice

After having discussed several historic telecommunications innovations,

we now turn to the analysis of a more recent case. Ethernet services have

been the subject of confl icting regulatory analysis and greatly diff ering

regulatory intervention. Some NRAs have deemed Ethernet services to

be a leased line and regulated them accordingly, whilst other have found

them to be in an ‘emerging market’ and therefore they have withheld from

any regulatory intervention.26 How could such large discrepancy arise?

We will show how the method just presented can be put to use to test

whether markets and services associated to an evolving technology require

special regulatory provisions, such as an emerging market connotation.

We propose that network innovation can lead to an emerging market only

where the risk of technological change is borne interdependently by more

than one player.

Ethernet, originally a local area network (LAN) technology, has had

a dramatic impact on the way in which data centers are built.27 It is

increasingly the pre- eminent means by which data equipment is interfaced

Internet and Applications

(WWW/email/etc)

Risks external to service provider

HighLow

Digitalization

Ris

ks

inte

rnal

to s

ervic

e p

rovid

er

Low

B-ISDN

Mobile

High

Optical transmission

Intelligent

Network

Figure 13.2 Positioning of historic examples


and devices are interconnected (it is now the norm for IT – information

technology – equipment to be supplied with an Ethernet socket). We will

show that the concept of Ethernet holds diff erent meanings. We will thus

use our model to analyze this pervasive technology and then draw policy

implications from this instructive case.

We need to keep in mind that Ethernet services (our focus) are a related

though distinct concept from Ethernet equipment. The latter follows the

Ethernet standards and interfaces and constitutes the switching equipment

in Ethernet networks. It is a very attractive technology for service provid-

ers as it has many useful technical features and is highly cost- eff ective.

However, somewhat ironically and counter- intuitively, current Ethernet

standards imply that Carrier Scale Ethernet services cannot be built upon

Ethernet equipment. Ethernet services are supplied by SPs which can

connect together the customer’s equipment using Ethernet interfaces. As a

result, Ethernet equipment and services have been and are still independ-

ent and mutually exclusive.

The development and deployment of Ethernet equipment initially had

no impact on telecom SPs who continued to supply leased lines and PSTN

calls with dial access. However, SPs also spotted a low- risk market oppor-

tunity: replace the traditional interfaces on the ends of their leased line

services with Ethernet interfaces, thus off ering Ethernet Leased Lines as

Ethernet Point-to- Point Services (point 1b in Figure 13.2). This change did

not result in any major technology shift for the SPs; the leased lines contin-

ued to be provided by the same existing underlying network.28 However,

the customer now had a service which was able to directly interface with

their own IT and data equipment. The only physical and practical diff er-

ence between the traditional wholesale leased line service and the Ethernet

leased line service comes from deploying Ethernet ‘boxes’ in place of their

equivalents (for example, Synchronous Digital Hierarchy, SDH) at the

customer site and at the point of interconnection between the incumbent

operator network and the entrant’s network.

Completely independently, SPs have also deployed Ethernet equipment

within their networks in order to interconnect equipment in a similar way

to end- users (SP LAN: point 2 in Figure 13.2). This deployment had no

impact ‘outside’ the sector and has been of limited technology risk to the

SPs themselves.29 However, the outlook for Ethernet technology suggests

that there may be room for more innovation. Ethernet equipment is a

highly cost- eff ective technology and there are standards developments

under the banner of Carrier Ethernet30 which can make this technology

useable by network operator for Ethernet services. Any decision to deploy

Carrier Ethernet is an internal- only decision with any mutual dependency

risk (point 3 in Figure 13.2).


As Figure 13.2 depicts, neither present- day Ethernet applications

(equipment or services) nor future developments present the condition

of interdependent risk which we argue needs to underlie an assessment of

emerging market. Hence, while (inadvertent) spillover eff ects aff ect how

the industry can benefi t from such technologies, the nature of the innova-

tion is not such as to require particular regulatory oversight.

Regulation, given its ex ante nature, cannot rely on retrospective obser-

vation to determine whether innovation constitutes an emerging market

(with the connotation of light- or ‘lighter’- touch regulation associated to

this). The introduction of Ethernet presentation leased line services (in

2005–07 for most European markets) is a clear example of this challenge.

In several countries, incumbents started off ering Ethernet services to their

retail customers, although usually without wholesale equivalents.31 NRAs

across Europe reacted diff erently.

Above in this chapter, we have tested our simple inside–outside model

in order to analyze risk in Ethernet- related technological change. Ethernet

has been used as a local area networking (LAN) technology since the

1970s and it is now the predominant technology in this area. In fact, an

SP introducing Ethernet services for leased lines is not exposed to interde-

pendent risk.32

In another context, network technologists are currently working on

extending Ethernet technology – Carrier Ethernet – for more direct use

within the network operator’s network. Whether this is successful, remains

to be seen; however, the risks are internal to the network operator. Our

simple model helps us shed light on what could be a more robust approach

to regulating markets aff ected by technologies such as Ethernet. In the fi rst

context, Ethernet should be regarded as part of generic technical progress

(which ironically emanated from outside the telecoms sector altogether)

and which had little or no deployment risk to telecom operators.33 In

this case, the innovation would fall into the category of service enhance-

ment and network upgrade (ERG, 2006). Thus no remedy arising from

Signifi cant Market Power (SMP) would have to be materially aff ected.

This example suggests that some general principles of market analysis

need to be developed in order to disentangle what are highly innovative

and risky activities due to the interdependent nature of risks (see Scenario

C in section 13.5.2 below) from those where development or deployment

are exclusively inside or outside the domain of any SP (Scenario B). Our

model can thus be used to facilitate a common approach when similar situ-

ations arise in the future.

In the case of Ethernet, a superfi cial analysis would suggest that this

technology can simultaneously present elements of risk both internal and

external to the action of a service provider wishing to introduce it. This


would support a classifi cation within an emerging market, thus hard to

grapple with by regulation. Instead, this section demonstrates that the

evolutionary path of Ethernet technology shows that such network inno-

vation presents risk which is either exclusively internal or external to the

action of SP. As a consequence this dispels the notion that the nature and

levels of risk warrant classifi cation of Ethernet Point- to- Point services as

an emerging market, while suggesting that these rest well within existing

regulated markets (see Figure 13.2).

13.5 CONCLUDING REMARKS: INNOVATION INTERDEPENDENCE AND MARKET REVIEWS

13.5.1 The Current Framework

It is commonly accepted that regulatory frameworks emphasizing the

application of the economic tools of competition law are suited to a mature

environment of stable outputs, costs and prices and indeed stable industry

structures. When, as is frequently the case, technological progress and

potentially disruptive technologies lead to economic change and disequi-

librium, these tools suff er from the lack of extensive information available

(Richards, 2006, 2007). In these circumstances, qualitative assessments,

like the simple model we have introduced, can become more important,

although they are limited by their judgmental nature.

European regulatory agencies are well aware of this dilemma and while

reviewing the application of remedies to established markets – as informed

by the European Commission (EC)’s guidance) – they have discussed the

issue of emerging markets (European Regulators Group, ERG, 2006). As

the ERG points out, a combined reading of the EU Framework Directive

and the European Commission’s SMP guidelines suggests that regulation

in emerging markets should not be ‘inappropriate’, hinting that emerg-

ing markets could be handled under competition law.34 In particular, the

Commission envisages a test to determine the emergent market status as

a ‘dynamic’ application of the Hypothetical Monopolist Test (HMT) on

both the demand and supply sides. Finally, the Commission argues that

the three- criteria test (TCT),35 which EU NRAs need to apply to defi ne

markets other than those specifi ed by the Commission, is the ultimate

test of whether a market is emerging or not. The ERG distances itself

from these assertions and policy recommendations, arguing that it is not

possible to apply the TCT under circumstances of immature demand and

supply. The ERG thus advises NRAs to engage in ‘close monitoring’

of situations where that is the case and stresses that a simple ‘network


upgrade’ is insuffi cient to aff ect the outcome of the TCT (ERG, 2006, p.

20).

While there are indeed good grounds for agreeing with the ERG that the

HMT and the TCT are largely superfl uous in a situation of dynamic change

and uncertainty, by the same token it is not evident what ‘close monitor-

ing’ defi nes. That regulation should not according to the Commission be

‘inappropriate’ is unhelpful to gain any regulatory certainty which may in

practice help fi rms in devising their conduct. From this guidance, NRAs

are still largely left in a policy vacuum if emerging markets can neither

be defi ned in principle nor be identifi ed in practice using the established

methodologies.

What we wish to emphasize is that similar regulatory discussions are

focused on innovation which is implicitly at the level of the single fi rm,

unlike much of network innovation. The latter, as we have highlighted

throughout our analysis, relies on multiple parties to upgrade simultane-

ously or at least to move broadly in compatible steps of an evolutionary

path. For instance, barriers to entry may not fall just on one party but

involve the activities of several players, each with its incentives to inno-

vate. This contrasts with the legacy of a focus on short- term assessments

of SMP on incumbent telecom operators which still informs the EU regu-

latory regime to a certain extent. We suggest that the key challenge for

regulatory policy- and lawmakers is to devise a framework which is able

to handle the strategic interplay of many parties over a much longer time

period.

13.5.2 Recommendations for Regulatory Policy

A study in a similar vein to our work proposes a taxonomy for emerging

markets in the telecommunications industry which centers on the underly-

ing uncertainty about the evolution of future demand (Crocioni, 2008).36

Our simple model complements this approach by suggesting a way to

break down the risk component in technological development which is

also focused on the supply side and which could be a fi rst step in regula-

tory assessments. We argue that there is considerable merit in operat-

ing the regulatory framework in such a way that it signals that diff erent

scenarios – based on the nature of innovation risk – will indeed be treated

diff erently in regulatory terms. Based on Figure 13.1:

Scenario A (bottom left quadrant): both the internal and exter- ●

nal components of risk are low. This is the case for existing but

mostly fully developed service (already largely with sunk investment

costs). Traditional quantitative methods can assess the scope for


appropriate intervention such as access regulation, quality of service

obligations, and so on.

Scenario B (top left and bottom right): only one of the risk compo- ●

nents is high. Here the decision as to whether and how to regulate

an ‘emerging service’ for which the technological characteristics are

well understood needs only to take special notice of whether the

pace of rollout may be aff ected by regulation. Similarly with respect

to possible further process innovation (insofar as investment costs

are still to be sunk).

Scenario C (top right): highly interdependent risk. Any regulatory ●

intervention cannot embody any meaningful stance toward invest-

ment in speculative and innovative new product or service research

and development (R&D) and subsequent deployment. Pre- emptive

regulatory forbearance can be a valid policy, subject to regular reas-

sessment informed by strategic analyses of technological change.

In all of the scenarios, the degree of complementarity in innovation will

have a major bearing on how risks are shared between diff erent parties and

the nature of those risks. The appropriate regulatory responses could be

quite diff erent depending upon the particular circumstances. Lacking this,

there will be a disincentive to commit to invest in innovatory activities and

infrastructures.

Analyses informed by this preliminary step could tackle more consist-

ently the following issues: whether to impose ex ante regulation and the

assessments of potential foreclosure; the relationship to any policy to

promote innovation as part of technology or industrial policies; and the

promotion and regulation of standardization and interoperability policies.

In general, our simple contribution could be a useful fi rst step to tackle the

public policy challenge of ascertaining ex ante whether market power is

likely to arise and which party or parties may hold it, in a scenario where

innovative eff orts are highly interdependent across organizations in the

communications industry.

NOTES

1. The views expressed in this chapter refl ect those of the authors alone and not necessarily those of the organizations to which they are affi liated.

2. In its turn, the discussion on the impact of competition on innovation is still open- ended. For instance, while some cite Arrow (1962) in stressing that perfectly com-petitive markets augment the incentives to innovate, others are more sympathetic to Schumpeter (1939, 1942), who in his later years highlighted the role of large corporate players on investing in wide- ranging innovation (the so- called ‘Mark II’ theory).


3. Capron Noll (1970) note: ‘Many services normally regarded as natural monopolies have remained natural monopolies only because of the interaction between regulation and technological change.’

4. In the EU, national regulatory authorities (NRAs) refer to guidelines issued by the European Commission’s Directorate General for Competition (European Commission, 2002) while performing the task of market analysis.

5. In the UK, Ofcom promoted the transition from the Network Interoperability Consultative Committee (NICC) technical standards committee (focussed on the technical terms of BT compliance) to the Next Generation Networks UK (NGNUK) forum, aiming to facilitate the advent of next generation networks (NGNs). NGNUK is ‘a coordination forum in which key investors in NGN infrastructure and services will discuss research, consider and, where possible, agree the direction for NGNs in the UK’ (www.ngnuk.org.uk).

6. This is due not only to the nature of distribution network (infrastructure) but also to access to content. In general, both direct and indirect network eff ects aff ect the provi-sion of communications services. See, for instance, Newberry (1999).

7. A real option can be associated to the managerial ability to respond to new information as it emerges; for example, to defer, to mothball or to exploit an unanticipated opportu-nity, and so on (Trigeorgis, 1995; Schwartz and Trigeorgis, 2004). Real options are not the only framework to assess risk. An alternative is the analytic approach of the capital asset pricing model (CAPM) which – due to its focus on the overall set of actions of a fi rm – may be not best placed to give appropriate insight into the risks associated with a specifi c decision to invest in a specifi c innovation.

8. The ITU is one of the oldest bodies in the United Nations system, having been founded in 1865.

9. Even in the automotive industry, some of the most radical innovations may require a systemic approach. For instance, hybrid vehicles need a distribution system which sup-ports electricity alongside petrol.

10. Note that standards are closely related to the transactions costs associated with asset specifi city. Often an outside supplier is only able to supply inputs to an adequate quality specifi cation if it makes ‘buyer- specifi c’ investment. Theoretically, this is not a problem if the investment has zero sunk cost (if there are effi cient second- hand markets for the equipment, and there are no signifi cant transaction costs associated with reselling). However, in practice, most investment in telecoms networks involves a signifi cant posi-tive sunk cost, in that future resale involves the fi rm losing considerable value. With positive sunk cost, suppliers are only likely to be prepared to make the buyer- specifi c investment if the buyer off ers a long- term contract which guarantees that the supplier can get an adequate return on the investment. However, even long- term contracting may not solve the ‘problem’. The point is that, once the supplier has sunk the invest-ment, this gives the buyer a stronger bargaining position, and an incentive to try to renegotiate more favourable terms. For a classic industrial organization study on this matter see Joskow (1987).

11. The precise technical specifi cation of some services can however be critical for certain applications and customer groups.

12. This is diff erent from the case of optical transmission innovation, which will be ana-lysed in the following sections.

13. ‘The requirement for Member States to ensure that national regulatory authorities take the utmost account of the desirability of making regulation technologically neutral, that is to say that it neither imposes nor discriminates in favour of the use of a particular type of technology, does not preclude the taking of proportionate steps to promote certain specifi c services where this is justifi ed, for example digital television as a means for increasing spectrum effi ciency’ (European Parliament, 2002; Framework Directive: Recital 18; emphasis added).

14. ‘The debate is not so much about whether to embrace technology neutrality as how much, and how soon. “Very few people disagree with the general concept of neutrality,”


says Mr Webb. It is, he says, a “motherhood and apple pie sort of thing” that nobody now wants to admit opposing’, The Economist (2005). Professor William Webb is Ofcom’s Head of R&D.

15. Studies of governmental interventions in Korea provide for several striking compari-sons with the EU landscape. See Kim et al. (2008), Jho (2007) and Picot and Wernick (2007).

16. The 1987 Green Paper analysed and assessed technology trends in order to devise the fi rst EC- wide telecommunications regulatory framework (European Commission, 1987).

17. This is evident from the very start of Article 8 of the Framework Directive, which sets out policy objectives and regulatory principles: ‘Member States shall ensure that in car-rying out the regulatory tasks specifi ed in this Directive and the Specifi c Directives, in particular those designed to ensure eff ective competition, national regulatory authori-ties take the utmost account of the desirability of making regulations technologically neutral . . . The national regulatory authorities shall promote competition in the provi-sion of electronic communications networks, electronic communications services and associated facilities and services by inter alia: . . . (b) ensuring that there is no distortion or restriction of competition in the electronic communications sector; (c) encourag-ing effi cient investment in infrastructure, and promoting innovation’ (European Parliament, 2002; Framework Directive: Art. 8.1, 8.2)

18. The European Commission acknowledged this long ago: ‘Regulators and market players alike face uncertainty as they look towards the future convergent environment. Regulators will need to have very clear objectives, including those of public interest, and a set of general purpose regulatory “tools”’ (European Commission, 1999, p. iv).

19. Capron and Noll (1970) note: ‘Many services normally regarded as natural monopolies have remained natural monopolies only because of the interaction between regulation and technological change.’

20. For instance, no direct references to the technology dimension or to innovation can be found among the policy objectives in the pivotal 1999 Communications Review (European Commission, 1999, p. iv).

21. For a detailed discussion on diff erent types of risk (in the context of next generation networks) see Ofcom (2005a, Section 9), where the nature and application of real option theories to regulation are debated.

22. For the sake of simplicity we are aggregating all external risks in the domain of several organizations other than our reference SP. This allows us a two- dimensional represen-tation; the model could be extended so as to include several axes which would lead to a more nuanced outcome, although not altering the key fi ndings.

23. Broadband Integrated Services Digital Network was the major programme of the ITU- T in the late 1980s and early 1990s.

24. The wide- ranging economic and societal consequences of such trends have been widely debated and are usually associated to the concept of ‘the death of distance’ (Cairncross, 1997).

25. Pre- optical transmission systems required expensive electronics every few kilometers to regenerate the signals. Thus, the overall network costs were mostly a function of distance and bandwidth. However, optical transmission reduced the need for regenera-tion equipment, previously often a dominant network cost. Many SPs did not change their price structures and still charged for services by distance and bandwidth, focusing on an ‘inside only’ view of the impact of this innovation. As a result customers still sought to optimize their demand to minimize the distance and bandwidth they needed to purchase and demand was predominantly for short, ‘thin’ connections. Networks thus confi gured had widely dispersed nodal electronics. Thus incremental cost analyses (such as those needed for price regulation) of such networks appeared still to show a strong dependence on distance and bandwidth. This was now arising from the nodal electronics demanded rather than the technological need for signal regeneration. This evolutionary path occurred especially in countries (or areas) with weak competition and


cost- oriented regulation based on historic network architectures. Where competition was unfettered and fi ercer, a forward- looking approach to pricing resulted in demand for longer, ‘fatter’ connections, aggregating traffi c onto fewer routes and generating a much more consolidated network structure. This evolution required a higher degree of industry awareness and coordination in order to address the element of mutually dependent risk for all parties. By doing so, industry players (fi rms, regulators) could overcome the prisoner’s dilemma nature of the situation and achieve a Pareto- superior outcome.

26. Some regulators have wavered in their policy on whether to include Ethernet amongst the leased lines market (which is subject to regulation). For instance the Spanish NRA, which had initially set Ethernet services apart from the regulated leased lines market, reversed its decision (CMT, 2006). A discussion of variations in regulatory stances towards Ethernet and other telecommunication services is in BT (2007, Annex 1).

27. In the early days of initial deployment, however, Ethernet competed with Token Ring and other LAN technologies (more recently with Fiber Distributed Data Interface – FDDI). For storage area networks (SANs) applications, Ethernet currently competes with Fibre Channel technology: it is yet unclear whether Ethernet will successfully supersede it in this fi eld.

28. This could be for instance synchronous optical networking/synchronous digital hier-archy (SONET/SDH), WDM, asymmetric transfer mode (ATM) or direct fi bre. More recently many SPs have exploited existing Multiprotocol Label Switching (MPLS) technology to off er Ethernet services.

29. There is however a commercial risk in that equipment may become redundant or lose its value more quickly than anticipated. In other words, old assets have to compete against new assets if the former are to remain in the market at all.

30. These in continuity fault management (CFM), provider backbone bridging (PBB), PBB – traffi c engineering (PBB- TE), and shortest path bridging (SPB).

31. This was not the case in the UK.32. This is because this technology uses the standardized equipment and interfaces of

LANs, but the ‘traffi c’ is actually still carried over the existing telecommunications SDH/ATM networks.

33. There could of course be potential fi nancial risk from devalued assets.34. This is also stated in the EC Wanadoo Interactive decision (European Commission,

2003).35. The three criteria are: high and non- transitory entry barriers; tendency towards eff ec-

tive competition; and insuffi ciency of competition law remedies. These are stated in Article 2 of the Commission recommendation on relevant product and service markets (European Commission, 2007).

36. Crocioni (2008) presents a model where uncertainty is augmented where: innovation and investment incentives are particularly important; or network eff ects (or switching costs) are relevant; or both of the above interact, adding complexity.

REFERENCES

Afuah, A. (2000). How much do your co- opetitors’ capabilities matter in the face of technological change? Strategic Management Journal, Special Issue: Strategic Networks, 21(3), pp. 387–404.

Arrow, K.J. (1962). Economic Welfare and the Allocation of Resources for Invention. The Rate and Direction of Inventive Activity: Economic and Social Factors. Princeton, NJ: Princeton University Press.

Baldwin, Carliss Y. and Kim B. Clark (2000). Design Rules: The Power of Modularity. Cambridge, MA: MIT Press.


Brandenburger, A.M. and B.J. Nalebuff (1996). Co- opetition. New York: Doubleday.

British Telecom (BT) (2007). The application of proportionate regulation in the electronic communications sector. In P. Richards and A. Tarrant (eds), The economic benefi ts from providing businesses with competitive electronic communications Services. Retrieved from http://www.btplc.com/Thegroup/RegulatoryandPublicaff airs/Consultativeresponses/BTdiscussionpapers/index.htm.

Cairncross, F. (1997). The Death of Distance. How the Communications Revolution will Change our Lives. Cambridge, MA: Harvard Business School Press.

Capron, W.M. and R.G. Noll (1970). Summary and conclusion. Technological Change in Regulated Industries. Washington, DC, Brookings Institution, pp. 197–238.

Commission del Mercado de las Telecomunicaciones (CMT) (2006). Resolucion por la que se aprueba la defi nición y análisis de los mercados de segmentos de terminación de líneas arrendadas al por mayor y segmentos troncales de líneas arrendadas al por mayor, la designación de los operadores con poder signifi ca-tivo de mercado y la imposición de obligaciones específi cas. Barcelona.

Crocioni, P. (2008). Leveraging of market power in emerging markets: a review of cases, literature, and a suggested framework. Journal of Competition Law and Economics, 4(2), 449–534.

Dobbs, I. and P. Richards (2004). Innovation and the new regulatory framework for electronic communications in the EU. European Competition Law Review, 25, 716–30.

The Economist (2005). Spectrum of opinion. 375(8428), 66.European Commission (EC) (1987). Towards a dynamic European economy:

Green Paper on the development of the common market for telecommunica-tions services and equipment. Brussels.

European Commission (EC) (1997). Towards an information society approach: Green Paper on the convergence of the telecommunications, media and infor-mation technology sectors, and the implications for regulation. Brussels.

European Commission (EC) (1999). The 1999 Communications Review. Towards a new framework for Electronic Communications infrastructure and associated services. Communication from the Commission to the European Parliament, the Council, the Economic and Social Committee and the Committee of the Regions. Brussels.

European Commission (EC) (2002). Commission guidelines on market analysis and the assessment of signifi cant market power under the community regulatory framework for electronic communications networks and services. Brussels.

European Commission (EC) (2003). Commission decision of 16 July 2003 relating to a proceeding under Article 82 of the EC Treaty (COMP/38.233 – Wanadoo Interactive). Brussels.

European Commission (EC) (2005). A European information society for growth and employment. Communication from the Commission to the European Parliament, the Council, the Economic and Social Committee and the Committee of the Regions. Brussels.

European Commission (EC) (2007). Commission recommendation on relevant product and service markets within the electronic communications sector sus-ceptible to ex ante regulation in accordance with Directive 2002/21/EC of the European Parliament and of the Council on a common Regulatory Framework


for electronic communications networks and services, 2007/879/EC, 28.12.2007. Brussels.

European Parliament (2002). Directive 2002/21/EC of the European Parliament and of the Council of 7 March 2002 on a common regulatory framework for electronic communications networks and services (‘Framework Directive’). Brussels.

European Regulators Group (ERG) (2006). Revised ERG Common Position on the approach to Appropriate remedies in the ECNS regulatory framework. Final Version May 2006. ERG 06 33.

Farrell, J. (2003). Integration and independent innovation on a network. American Economic Review, 93(2), 420–24.

Farrell, J. and P.J. Weiser (2003). Modularity, vertical integration, and open access policies: towards a convergence of antitrust and regulation in the Internet age. Harvard Journal of Law and Technology, 17(1), 85–134.

Fransman, M. (2007). The New ICT Ecosystem. Implications for Europe. Edinburgh: Kokoro.

Gawer, A. and R. Henderson (2007). Platform owner entry and innovation in complementary markets: evidence from Intel. Journal of Economics and Management Strategy, 16(1), 1–34.

Gunasekaran, V. and F.C. Harmantzis (2008). Towards a Wi- Fi ecosystem: tech-nology integration and emerging service models. Telecommunications Policy, 32(3–4): 163–81.

Henderson, Rebecca M. and Kim B. Clark (1990). Architectural innovation: the reconfi guration of existing product technologies and the failure of established fi rms. Administrative Science Quarterly, 35(1), 9–30.

Jho, W. (2007). Global political economy of technology standardization: a case of the Korean mobile telecommunications market. Telecommunications Policy, 31(2), 124–38.

Joskow, P.L. (1987). Contract duration and relationship- specifi c investments: empirical evidence from coal markets. American Economic Association, 77, 168–85.

Kim, Y., H. Jeon and S. Bae (2008). Innovation patterns and policy implications of Adsl penetration in Korea: a case study. Telecommunications Policy, 32, 307–25.

Knight, F.H. (1921). Risk, Uncertainty and Profi t. New York: Houghton Miffl in.

Langlois, N. (2002). Modularity in technology and organization. Journal of Economic Behavior and Organization, 49(1), 19–37.

Majone, G. (1996). Regulating Europe. London: Routledge.Newberry, D. (1999). Privatization, Restructuring and Regulation of Network

Utilities. Cambridge, MA: MIT Press.Offi ce of Communications (Ofcom) (2005a). Ofcom’s approach to risk in the

assessment of the cost of capital. Final Statement. London.Picot, A. and C. Wernick (2007). The role of government in broadband access.

Telecommunications Policy, 31(10–11), 660–74.Richards, P. (2006). The limitations of market- based regulation of the electronic

communications sector. Telecommunications Policy, 30 April, 201–23.Richards, P. (2007). Technical progress, market evolution and the regulation of

the ECS in the EU. Competition and Regulation in Network Industries, 8(2), 165–257.


Robertson, L. and N. Langlois (1995). Innovation, networks, and vertical integra-tion. Research Policy, 24(4), 543–62.

Schumpeter, J.A. (1939). Business Cycles: A Theoretical, Historical, and Statistical Analysis of the Capitalist Process. New York: McGraw Hill.

Schumpeter, J.A. (1942). Capitalism, Socialism and Democracy. New York: Harper & Brothers.

Schwartz, E.S. and L. Trigeorgis (2004). Real Options and Investment under Uncertainty: Classical Readings and Recent Contributions. Cambridge, MA: MIT Press.

Trigeorgis, L. (1995). Real Options in Capital Investment: Models, Strategies, and Applications. Westport, CT: Praeger.

301

14. Next generation mobile networks deployment and regulation in the European Union1

Claudio Feijóo, Sergio Ramos and José- Luis Gómez- Barroso

14.1 INTRODUCTION

The ubiquitous broadband revolution will fundamentally change the land-

scape of European telecommunications. Citing European Commissioner

Reding (2008): ‘we are currently confronted with a once in a generation

opportunity to make sure that Europe promotes and leads the next phase

of wireless technology which will be the transition from voice and short

text services to the wireless web’. Today mobile telecommunications are a

major driver of the European Union (EU) economy, with expected market

growth by 2.4 percent in 2009 to reach an overall value of €144.5 billion.

Mobile data traffi c, including text messages and Internet access, contrib-

utes for €33 billion of the total turnover in mobile telecommunications and

most of the growth, 8.4 percent (EITO, 2009).

The so- called next generation networks (NGNs) will be the supporting

infrastructure of ubiquitous broadband. They are typically defi ned as mul-

tiservice networks, running over Internet Protocol (IP)- based networks,

complemented by fl exible service platforms and management systems.2 In

particular they will be able to make use of multiple broadband, quality of

service (QoS)- enabled transport technologies and in which service- related

functions are independent from underlying transport technologies. They

will also off er access by users to diff erent service providers and support

generalized mobility which will allow consistent and ubiquitous provision

of services to users (Kocan et al., 2002; Modarressi and Mohan, 2000).

For the purposes of this chapter, an NGN will be simply a single network

which delivers multiple applications – voice, data and video – to multiple

devices, whether fi xed or mobile.

The conditions for the deployment of the access part of NGNs are

currently the key topic in the review of the electronic communications


regulatory model in the EU (Cave and Picot, 2008). The main issues raised

during the debate are the conditions for the return on investments, the

type and level of competition and the model required to promote innova-

tion. The subset of NGNs which specifi cally will support the evolution

from today’s mobile communications infrastructures are usually called

next generation mobile networks (NGMNs). They confront basically the

same regulatory issues as NGNs. But they present, in addition, some par-

ticularities of interest: the need for adequate spectrum, the evolutionary

nature of mobile technologies, and their distinctive regulatory model to

date (Ramos et al., 2004).

This chapter’s aim is to examine the impact of regulation in the deploy-

ment of NGMNs. With that objective the next section introduces briefl y

the concept of NGMNs and the overall status of the market. From here the

examination of the regulatory elements that infl uence the deployment of

NGMNs follows and, fi nally, the chapter closes with some conclusions.

14.2 THE STATUS OF THE NGMNS

NGMNs are regarded as a future platform for ubiquitous broadband,

facilitating the smooth migration from existing mobile infrastructures,

and allowing for the commercial launch of new mobile services and appli-

cations. Thus, in the mobile carriers’ vision, NGMNs will ensure a virtu-

ous cycle of investment, innovation and adoption of services (NGMN

Alliance, 2006). From a strict technological perspective NGMNs encom-

passes fourth generation, 4G- type mobile technologies, such as Long

Term Evolution (LTE), Mobile WiMAX (Worldwide Interoperability for

Microwave Access), IMT- Advanced3 (Ring, 2008) and convergent technol-

ogies like femtocells, wireless technologies – Near Field Communication,

NFC, for instance – completing a ‘network of sensors and tags located on

surrounding objects’ (Schwarz- da- Silva, 2008).

Table 14.1 overviews some of the main milestones and features of

the technologies needed for the deployment of NGMN. Furthermore, a

number of experts from the industry forecast that mobile broadband con-

nections will overcome fi xed ones sometime around 2011–13 (Nerandzic,

2008; Ouvrier, 2008). As an obvious result from these ambitious road-

maps, mobile networks will require access to new spectrum and/or a much

more effi cient management of it. Without these spectrum improvements it

will be impossible to deploy rapidly enough the required technologies that

satisfy users’ demands or to compete – and complement – satisfactorily

with fi xed broadband technologies.

In fact the deployment of true NGMNs, albeit slowly, has already

303

Table

14.1

S

um

mary

of

tech

nolo

gic

al

roadm

aps

for

NG

MN

dep

loym

ent

HD

PA

(3

.5 G

)L

TE

(4

G)

Fem

toce

lls

Mo

bil

e

WiM

AX

IMT

Ad

va

nce

dN

FC

Th

eore

tica

l

maxim

um

data

rate

s

14

Mb

ps

(do

wn

stre

am

)

40

Mb

ps

(do

wn

stre

am

in H

DP

A+

)

5.6

Mb

ps

(up

stre

am

)

10

0 M

bp

s

(do

wn

stre

am

)

50

Mb

ps

(up

stre

am

)

–

50

Mb

ps

1 G

bp

s1 M

bp

s

Typ

ical

data

rate

s

3.6

Mb

ps

(do

wn

stre

am

)

2 M

bp

s

(up

stre

am

)

––

1

0 M

bp

s1

00

Mb

ps

400 K

bp

s

Ch

an

nel

ban

dw

idth

5 M

Hz

1.2

5–

20

MH

z–

1

.25

–2

0 M

Hz

Up

to

10

0 M

Hz

2 M

Hz

Beg

in o

f m

ass

ive

dep

loym

ent

20

08

–1

02

00

9–

12

20

09

–1

0

20

10

–1

1 (

for

ha

nd

sets

)

20

09

–1

12

01

3–

17

2011–17

Cri

tica

l

tech

no

log

ies

MIM

OO

FD

MA

Ma

na

gem

ent

OF

DM

AD

yn

am

ic s

pec

tru

m

ma

na

gem

ent

Sec

uri

ty

Main

ad

van

tages

Ev

olu

tio

na

ry

fro

m e

xis

tin

g

3G

Ev

olu

tio

na

ry f

rom

3.5

G

Fix

ed- m

ob

ile

con

ver

gen

ce

Incr

ease

of

cov

era

ge

No

t a

leg

acy

tech

no

log

y

Ev

olu

tio

na

ry f

rom

4G

Sm

art

en

vir

on

men

t

Main

dis

ad

van

tag

es

Tra

nsi

tio

n

tech

no

log

y

Tim

e- to

- ma

rket

Inte

gra

tio

n

in e

xis

tin

g

net

wo

rks

Bu

sin

ess

case

for

new

tech

no

log

y

Sti

ll i

n e

arl

y s

tag

es

of

sta

nd

ard

iza

tio

n

pro

cess

Bu

sin

ess

case

fo

r

dep

loym

ent


begun. A survey from the Global mobile Suppliers Association (GSA)

in September 2007 stated that 138 high- speed downlink packet access

(HSDPA) networks had been commercially launched services in 66 coun-

tries, compared to 58 commercial launches in 37 countries the previous

year. By February 2008 125 HSDPA network operators had commercially

launched or were deploying 3.6 Mbps peak or higher downlink data rate

capability. At the end of 2008 the number of commercial HSDPA deploy-

ments had increased to 237 in 105 countries, and 167 commercial HSDPA

networks supported 3.6 Mbps or more, including 81 which supported 7.2

Mbps peak or higher downlink speeds. The latest data (GSA, 2009) show

that several operators are launching HSPA+ in 2009 for at least 21 Mbps

peak downlink speeds and that there are over 20 mobile operators already

committed to deploy LTE.

However, availability of mobile broadband is only one condition

needed to reach mass adoption. Aff ordability of mobile broadband con-

nections, utility and appeal of new mobile applications and services, as

well as the availability and the usability of new mobile handsets are also

required (Feijóo et al., forthcoming). For these reasons, mobile data serv-

ices continue to rely on simpler services like messaging which accounted

for around 14 percent of total revenues of mobile operators in 2007. While

more advanced data services have shown signifi cant growth recently, they

still accounted for just around 7 percent of total revenue compared to 5

percent in 2006. The only two major exceptions in 2007 were Japan and

Korea, where mobile content average revenue per user (ARPU) was 30.5

percent and 18.9 percent respectively (Netsize, 2008).

No account of the future status of NGMNs can be complete without

an examination of the demand side, especially in view of the increasingly

important role of users in the so- called mobile 2.0 (Jaokar, 2006). Figure

14.1 shows the forecast for the evolution of mobile social networking users

(in millions) compared with those of total mobile subscribers, users of 2G,

beyond third generation (B3G) technologies and mobile Internet.4

In summary, the mobile sector, particularly in Europe, is at a turning

point. As the revenues from voice remain fl at or even decline (EC, 2008),

there is the need to go further to compete with fi xed players on broadband

access, to off er new attractive mobile services, applications and content

and fi nd fresh business models for them (Ramos et al., 2009).

14.3 REGULATION AND NGMN

From a regulatory perspective, the deployment of NGMN is linked with

two main pillars of the electronic communications framework: spectrum

Next generation mobile networks in the European Union 305

management and the triangle ‘investment – competition – innovation’.

Together, they form what we could name as ‘the model’ for the future

mobile industries from the public perspective. Both of them are currently

under review due precisely to their impact on the deployment of NGNs.

14.3.1 Spectrum Management

NGMNs will need new spectrum if a truly ubiquitous broadband con-

nectivity is to be achieved in a competitive environment. The current

allocation of spectrum to mobile operators will not be enough to satisfy

the roadmaps shown in Table 14.1. Even counting on technology devel-

opments to increase the effi ciency (bits per Hertz) of radio transmission

techniques, the implementation of 4G- type mobile communications will

require from 20 to 100 MHz of spectrum per operator (in a design of cel-

lular coverage with high quality of service for high data throughputs).

The total budget of spectrum for NGMNs will depend on the availability

of spectrum, on the level of competition (that is, the number of mobile

network operators in the market5) and on the expected return on invest-

ments. Higher quality of service will need more spectrums per user.

The fi rst of these factors, radio spectrum availability, is linked with an

effi cient spectrum usage and management (Feijóo et al., 2009a). In fact,

6000

5000

4000

3000

2000

2007 2008 2009 2010 2011 2012 2013

1000

0

Total subscribers

2G users

Mobile Internet users

B3G users

Mobile social networking users

Source: Own calculations from industry data.

Figure 14.1 Evolution of world mobile subscribers (millions) of diff erent

technologies and services: 2G, B3G, mobile Internet and

mobile social networking


the traditional model for spectrum management appears as excessively

complex and lacking fl exibility, which has led to the reinforcement of a

current opinion favorable to the introduction of market economy criteria

in radio electric spectrum management, an idea already argued by Coase

(1959). This general principle of effi ciency can be broken down into three

dimensions: technical effi ciency, economic effi ciency and social effi ciency.

Technical effi ciency implies that, should there be several technologies to

off er a specifi c service under the same conditions, the one that uses the

least resources – frequencies – should be the preferred option. An eco-

nomically effi cient spectrum allocation is one in which there is no way

to reallocate spectrum such that everyone is better off , everything else

being equal (Morris, 2005). Social effi ciency implies that the radioelectric

spectrum should be used in those services or applications maximizing the

well- being of society. In all cases, the spectrum effi ciency analysis should

be extended by introducing a prediction of future development or, in other

words, optimizing a dynamic defi nition and not a simply static one.

In fact, many national authorities in charge of spectrum management,

the European Commission or the Internet Telecommunication Union

(ITU),6 have released documents supporting the modifi cation of the

current spectrum management models and defending the inclusion of

market mechanisms. Therefore, that the radioelectric spectrum needs a

new management model – or at least a renewed one – is an assertion over

which there exists a consensus among public administrations, players

with interest in the market and researchers. The new ways of manag-

ing the spectrum intend to increase fl exibility and transparency, as well

as the speed of the response to technological innovations. The actual

demand and the value given by the market to each band of frequencies

are new governing criteria. Obviously, any change should improve global

resource usage effi ciency, while maintaining certain inalienable technical

requirements.

There are three possible measures7 leading to change. These are listed

in ascending order in the process of assignment of the spectrum usage

rights:

Modifi cations in the conditions of the license: relaxation of some ●

clauses but, especially, authorization of the transmission of rights

(secondary spectrum trade).

Modifi cations in the license- assignment mechanisms: usage of ●

auctions.

Modifi cations in the defi nition of the licenses: avoiding binding the ●

license to specifi c technologies – technological neutrality; or even to

specifi c services – service neutrality.


Although this chapter’s aim is not the discussion about the merits of the

reform of spectrum management but its relationship with NGMNs, it is

worth pointing out that some of these measures are already in use. Secondary

spectrum trading has been adopted by the responsible bodies of some

member countries (Ofcom, 2008b), although in general only as pilot projects

or in frequency bands with moderate economic value and reduced capacity

for provoking interferences. The results of spectrum trade so far are very

modest (Ofcom, 2008a), but due to existing limitations they are inconclusive

about the future behavior of the spectrum market in diff erent circumstances.

About spectrum auctions it can be said that they are increasingly used as a

more transparent procedure to allocate frequencies to players; however,

‘beauty contests’8 are still a popular mechanism for relevant frequency bands

across EU member countries. Again, the objective of the chapter is not to

defend auctions as opposed to beauty contests (see Bourdeau de Fontenay,

1999; Cramton, 2002) but to highlight that the allocation and assignment

procedures are not harmonized at all. Last, technological neutrality in spec-

trum has been the subject of several recent ‘refarmings’ of spectrum, mainly

to allow the use of 3G – and beyond – standards for mobile communications

in bands originally allocated exclusively for 2G. In general, service neutrality

is not yet in place for any relevant spectrum- based service.

In any case the application of these reforms for the deployment of

NGMNs leads to a number of short- term opportunities, being the most

prominent the so- called ‘digital dividend’ resulting from the television

‘analog switch- off ’. Digital dividend, located in the VHF–UHF (very high

frequency, ultra high frequency) bands, is the ‘beachfront spectrum’ (so-

called in the USA due to its techno- economic features). The cost of the

deployment of a cellular technology is directly proportional to the number

of base stations required (Giles et al., 2004). The use of the 700 MHz band,

for example, reduces about seven times the number of base stations com-

pared with the 3.5 GHz band for the same coverage conditions (Forge et

al., 2007). This reason, together with better indoor coverage and simpler

technology, are enough to explain the interest in these classes of frequen-

cies. Nevertheless, if we take into consideration the cost per user at a given

quality of service, it is interesting to note that the use of lower frequencies

– and, therefore, larger cell sizes – asks for larger spectrum bandwidths to

maintain the quality. Alternatively, ‘macrocells’ are only practically appli-

cable in areas – rural, for instance – where the number of concurrent users

does not increase enough to set operations beyond the point of quality

degradation (Johansson et al., 2004).

The problem lies, obviously, in the fact that this highly desirable spec-

trum is already allocated, mostly to the television broadcasters. However,

as mentioned, the transition to the digital terrestrial television opens a


window of opportunity to a refarm of the spectrum, and to change some

of the rules of the game in spectrum, management as mentioned above. In

practice this process is not smooth and it is a good example of the diffi cul-

ties to achieve a solution accepted by all parties, in spite of the European

Commission (EC) eff orts for the adoption of a common perspective on the

use of spectrum available from the digital switchover (EC, 2007).

The following two positions illustrate the problem. Forge et al. (2007) in

a study for T- Mobile about the use of digital dividend in the EU conclude

that:

estimates of the accumulated eff ect indicate as much as an additional 0.6% GDP growth per year for the EU economy by 2020 in the mobile case when compared with broadcast television . . . allocating spectrum to enable wire-less broadband could have a dramatic impact on bridging the digital divide by using the new spectrum to provide access for all across the EU’s 27 member countries.

It is worth noting that this, sometimes forgotten, important eff ect on the

digital divide is also supported by the other authors, like Cave and Hatta

(2008) and Gómez- Barroso and Robles- Rovalo (2008). On the opposite

side, a ‘counter report’ supported by the European Broadcasting Union

(EBU) (Oliver and Ohlbaum Associates, 2008) argues that:

a market for UHF spectrum is likely to fail and not allocate a socially optimal amount of spectrum to terrestrial television. This is because terrestrial TV gen-erates signifi cant public value for society that would not be visible in any hypo-thetical contest for spectrum with other uses, and cannot easily be replicated through provision of TV using other platforms. The medium- term value that could be created by other uses of UHF spectrum, including rural broadband – appears modest.

Reading these lines, it is easy to realize that the endeavor of a pan-

European harmonization of the digital dividend in search for the appro-

priate economies of scale currently has to confront high barriers. A debate

so bitter has, however, several possible solutions: on the one hand, there is

spectrum enough, if not to please every player, at least to fi nd a reasonable

compromise,9 and on the other hand, it is possible just to let the evolution

of services – addition of interactivity to the broadcasting side or use of

audiovisual on the communications side – dissolve the diff erences between

them. As an argument for the fi rst option, it is worth recalling that in the

USA, where there is certainly no need to harmonize across member states,

the auction for the digital dividend has already taken place in 2008, even

including some conditions for ‘openness’ – still to be seen in its practical

implementation – in some of the frequency blocks (FCC, 2008).


Apart from digital dividend there are more examples of immediate

infl uence of spectrum management in NGMN deployment. For instance,

although as mentioned earlier the cost of cellular systems is proportional

to the number of base stations, it would be possible to arrive at a diff erent

effi cient solution decreasing the cost of each of them to compensate for the

increment in their number.10 The resulting ‘pico’ or ‘femto’ cells will require

a fair amount of neutrality in spectrum usage to adapt them continuously

to the changing landscape of mobile and wireless communications tech-

nologies, business models and users’ demands. ‘Ad hoc’ mobile networks

would be another example of much- needed fl exibility in spectrum manage-

ment. Without exhausting the subject, one more example would be sharing

spectrum among operators to decrease deployment costs. This could be a

result of an ‘extended’ secondary spectrum trade.

Finally, in addition to making more fl exible and decreasing the costs

of NGMN deployment, it is worth mentioning that an effi cient spectrum

framework would increment the number of potential mobile network oper-

ators in the market and the overall level of competition in the electronic

communications domain (Cave and Picot, 2008). Equivalently, given a

certain amount of available frequencies, spectrum management will deter-

mine the initial level of competition and the rules for its evolution.

14.3.2 The Mobile Communications Framework: Investment –

Competition – Innovation

Investments in NGMNs are arguably the fi rst issue on the agenda of

mobile operators from the perspective of a review of the electronic

communications regulatory framework. In this regard, since its market

started in the early 1990s the EU mobile sector has enjoyed a mostly non-

intrusive regulatory model that has led it to be a reference worldwide. As

a consequence, the mobile business is characterized by operators’ leading

positions, controlling as many elements within the value chain as possible

(Sabat, 2002). The result from the perspective of mobile operators was the

well- known ‘walled garden’, ‘silo’ or ‘on- portal’ model, where content and

applications revenues are generated by operators within their own value

structure and where users are guided to stay confi ned as much as pos-

sible within this structure. The bottom- end rationale for a walled garden

in mobile content is the use of a scarce and costly resource: the mobile

networks.

Undoubtedly, in Europe this model had enabled competitive infra-

structure development with quite successful results until the migration

towards 3G networks started in 2000 (Ramos, 2005). From that point on,

the mobile business sector experienced diffi culties that forced operators to


delay 3G commercial launches by, on average, three to four years (Bohlin

et al., 2003). As a consequence during that time mobile players have had

to focus more on mobile value- added services than on broadband capacity

and, therefore, were not generally considered real broadband players until

very recently.

NGMN deployment changes the landscape dramatically. On one side

they require considerable investments (De- Antonio et al., 2006; Forge et

al., 2005) and therefore mobile operators argue for maintaining the status

quo or even for a relaxation of the competition rules – sharing infrastruc-

tures or integrating with fi xed operators, for instance – or, alternatively,

for special conditions – risk premiums – for investors. But on the other

side mobile operators also increasingly compete in a convergent scenario

with all types of operators. NGNs, and NGMNs as a part of them, will be

the fi rst major infrastructure investment for the overall sector to take place

in a relatively open competitive scenario. Experience, both for operators

and regulators, is scarce.

This fi xed–mobile convergence over NGNs will create a stimulated

level of competition where now both fi xed and mobile operators will

contend for the same ubiquitous broadband market.11 From the regula-

tory perspective it is not diffi cult to realize that this type of competition

will require a more convergent regulatory framework. Some symptoms

of it are already surfacing in the proposals for the next regulatory frame-

work for electronic communications in the EU,12 in the review of the

universal service obligations,13 in the major examples of the regulation of

mobile voice and short message system (SMS) roaming charges, or in the

reconsideration of termination charges and interconnection models in the

mobile domain. However, it is still unknown whether the approach will

resemble the heavy- handed regulation of the incumbent fi xed networks

or the light touch until now typical of the mobile domain. All in all, the

expected consequence of the converged competition would be a more level

playing fi eld for operators, independently of the technology they use.

More recently, innovation, the third relevant element in the mobile

communications policy framework, has re- emerged with strength. There

are two main reasons for it: the opportunities for growth and jobs derived

from mobile evolution, and the users’ increasing demand for an unre-

stricted and wide choice of content and applications, including their new

role in innovation.

As a consequence the traditional ‘walled garden’ approach of mobile

operators is fi nally eroding and slowly giving way to the opposite model

where the mobile operator is a mere provider of connectivity (Ramos et

al., 2002). In between, there are a myriad hybrid models with content and

application providers, enablers and brokers sharing revenues with mobile


operators (Feijóo et al., 2009b). However, the speed of change seems

not to be enough. Citing Holden (2008): ‘the level of control exerted by

[mobile] operators rankles with, and exasperates, the [application and]

content providers, an environment not necessarily conductive for the

introduction and mass adoption of innovative mobile services’. Openness,

resembling the successful ‘open garden’ Web 2.0 model of Internet, seems

to be the increasingly sought response to incentivize innovation and new

business models. What type of openness, how to enforce it and the reac-

tion of players are some of the issues remaining unsolved.

As a summary of the subsection, three major types of uncertainties

persist in the mobile communications regulatory framework with regard

to NGMN deployment: the conditions for investment in NGMNs (that is,

should regulation be the same as before? Should spectrum management

be used to ease the deployment?); second, the conditions for competition

(that is, should there be a fi xed mobile convergence in terms of competi-

tion and regulation?); and third, the conditions for innovation (that is,

which is the structure of the value- added chain providing more social

benefi ts in the long run?).

14.4 CONCLUSION

Next generation networks will support a renewed electronic communi-

cation structure where the market opportunities lie in the provision of

ubiquitous broadband connectivity and in the combination of services,

applications and content on top of this (Feijóo et al., forthcoming).

Mobile operators will deploy the mobile fl avor of NGNs to compete

among themselves and with fi xed operators. In practical terms, NGMNs

should provide wide coverage for at least 50 Mbps downstream in the 2015

horizon (Ramos et al., 2009).

However, the deployment of NGMNs by mobile operators confronts

hurdles both in the technological domain – uncertainties in the technol-

ogy roadmaps; and in the market domain – uncertainties about profi table

business models and the response of users. Regulation will not stop – or

start – the process of deployment of NGMNs, since the rationale behind

its deployment resides in the demands and expectations of the users. But

regulation, and from a complementary perspective industrial policy, will

impact the pace of such a deployment (Cave and Picot, 2008). It can

provide a stable and coherent framework for the market to evolve; or the

opposite: it could amplify the eff ect of these uncertainties, delaying the

deployment of NGMN.

The most immediate eff ect of regulation for NGMNs refers to spectrum.


Future – almost present – mobile or wireless access technologies will need

more effi cient access to broader- spectrum bands to be able to compete in

the ubiquitous broadband landscape. Therefore, the reform of spectrum

management has a prime role. There are three main conditions for such a

reform: (1) more fl exibility to accommodate the rapidly changing nature of

mobile technologies; (2) a search of techno- economic effi ciency to increase

the utility of a limited resource like spectrum; and (3) further harmoniza-

tion in spectrum management mechanisms with the double goal of avoid-

ing new digital divides and increasing the economies of scale to ease the

deployment in an already costly scenario. In the authors’ view, the main

conclusion from the spectrum management discussion is that its potential

benefi ts for NGMN deployment will not be evident until each and every

of the measures for reform are included in a comprehensive and coherent

EU- wide framework. It would not make much sense to have secondary

spectrum trading in place if the conditions to access such spectrum diff er

across countries or if there is not enough fl exibility for the use of it.

Precisely, this is the main diff erence when comparing EU with other

regions (for example, the US): the pressing need for harmonization.

With the background of the huge EU success of Global System for

Mobile Communications (GSM) standardization (and the initial US

delay in mobile communications derived from the many PCS solutions),

and the subsequent (and not less huge) failure of the Universal Mobile

Telecommunications System (UMTS), the authors are arguing for a diff er-

ent and subtler type of harmonization: it will be about the new conditions

for use of spectrum, not the actual technologies using it.

Independently of any of these circumstances, and extrapolating the

current pace of network deployment (and user adoption), 2010 and 2011

take shape as critical moments when the spectrum management deci-

sions should be in place. Therefore, the ITU’s World Radio Conference

(WRC) of 2011 arises as a major milestone in the process leading to 4G

and beyond; a date even more signifi cant when the extremely slow process

of spectrum management and standardization is acknowledged (Ring,

2008).

Spectrum management, however relevant, is only one part of the equa-

tion. It should be framed in a larger perspective: policy- makers and regula-

tors should defi ne a scenario for NGMN deployment as the cornerstone to

stimulate innovation that is transferred to markets and users in the form of

new services, applications and businesses. There are two minimum condi-

tions for this scenario to be achieved: (1) stability, since the time required

for the return on investments for the deployment of NGMNs will be

longer than for previous generations of mobile communications; and (2)

coherence, in the sense of the defi nition of a model – comprised policies on


spectrum management, investment, innovation and competition in NGN

– for mobile operators. Beyond these minimum conditions there could

be more ambitious and challenging objectives. These additional objec-

tives are related not to the NGMN deployment as such, but to the still to

be fully understood innovation which could originate from a ubiquitous

mobile broadband infrastructure. The new mobile innovation would

require: (1) a more open ecosystem to decrease the barriers of an hetero-

geneous and fragmented ecosystem in particular when compared with the

Internet (Feijóo et al., 2009b); (2) making the most of the new role of users

(Pascu, 2008); and (3) reviewing the institutional framework that supports

it (Fransman, 2007), including some additional pieces of the regulatory

framework like the intermittently debated convergence with audiovisual.

The fi nal challenge consists of meeting these goals while the mobile opera-

tors are still in the process of investing in NGMNs.

In conclusion, both elements of reform – spectrum management and

regulatory policy framework – have been on the agenda since the stage for

the initial liberalization of electronic communications was – approximately

– met in about 2002–03 in Europe, though they have never reached actual

and complete implementation. In the authors’ opinion, the window of

opportunity for mobile operators’ deployment of NGMNs is and will be

open up to the medium term and, to a great extent, it will be independent

of regulation as the LTE deployment commitments of operators have

proved even during times of economic crisis. However, the regulatory

policy framework severely impacts the intensity and speed of the NGMN

deployment and, through the ubiquitous broadband this will provide, of a

long- awaited new wave of applications and societal empowerment.

NOTES

1. The views expressed are purely those of the authors and may not in any circumstances be regarded as stating an offi cial position of their Institutions

2. The NGN ‘Working defi nition’ can be found at: http://www.itu.int/ITU- T/study-groups/com13/ngn2004/working_defi nition.html

3. IMT- Advanced refers to international Mobile Telecommunications Advanced set of standards for 4G type of mobile communications as defi ned by ITU- R, Note that although LTE is usually branded as 4G its fi rst release does not fully match the IMT- Advanced specifi cations.

4. Based on own calculations from market analysts’ data. Mobile 2.0 is foreseen to reach the impressive fi gure of 1000 million mobile social networking users around 2014.

5. In the EU- 27, the average – mode – number of mobile operators per member state is four (EC, 2008).

6. As an example, refer to the ITU document ‘Market mechanisms for spectrum manage-ment’, http://www.itu.int/osg/spu/stn/spectrum/workshop_proceedings/STN.MMSM- 2007- PDF- E.pdf.


7. Together with these measures, a series of general guidelines has been proposed for the practical introduction of reforms in the spectrum management mechanisms. The inter-ested reader can consult Feijóo et al. (2009a)

8. In a ‘beauty contest’ licenses are assigned according to the evaluation of a number of technical and economic criteria, typically including anything from type of technology used, deployment of coverage proposed and national added value and employment created to national innovation push.

9. The European Commission already proposes one such possible compromise (EC, 2007).

10. Therefore, in practice, mobile operators will require to be cost- eff ective in a range of cell sizes (macro, meso, pico, femto) and appropriate allocations of frequency bands and bandwidths (Johansson et al., 2004).

11. Note that integrated operators, that is, operators both with fi xed and mobile technolo-gies, are anticipating this trend and they position themselves in this new competition landscape.

12. See http://ec.europa.eu/information_society/policy/ecomm/tomorrow/index_en.htm.13. Communications from the Commission to the European Parliament, the Council, the

European Economic and Social Committee and the Committee of the Regions on the second periodic review of the scope of universal service in electronic communications networks and services in accordance with Article 15 of Directive 2002/22/EC, Brussels, 25.9.2008, COM(2008) 572 fi nal C.F.R. (2008).

REFERENCES

Bohlin, E., J. Björkdahl, S. Lindmark, T. Dunnewijk, N. Hmimda, S. Hultén and P. Tang (2003). Prospects for third generation mobile systems. Report No. EUR20772EN. Seville: Institute of Prospective Technological Studies.

Bourdeau de Fontenay, A. (1999). Auctions vs. beauty contests, is it the ques-tion? A new look at access and spectrum allocation in France and in the US. Communications and Strategies, 36, 111–23.

Cave, M. and K. Hatta (2008). Universal service obligations and spectrum policy. info, 10(5–6), 59–69.

Cave, M. and A. Picot (2008). Workshop next (‘now’) generation access (NGA): How to adapt the electronic communications framework to foster investment and promote competition for the benefi t of consumers? Summary, briefi ng notes and presentations (No. IP/A/ITRE/WS/2008–09 PE 408.554). Brussels: Department Economic and Scientifi c Policy, European Parliament.

Coase, R. (1959). The Federal Communication Commission. Journal of Law and Economics, 2, 1–40.

Cramton, P. (2002). Spectrum auctions. In M. Cave, S. Majumdar and I. Vogelsang (eds), Handbook of Telecommunications Economics (pp. 605–39). Amsterdam: Elsevier Science B.V.

De- Antonio, J., C. Feijóo, J.L. Gómez- Barroso, D. Rojo and A. Marín (2006). A European perspective on the deployment of next generation networks. Journal of the Communications Network, 5(2), 47–55.

EC (2007). Reaping the full benefi ts of the digital dividend in Europe: a common approach to the use of the spectrum released by the digital switchover. Retrieved from http://eur- lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:52007DC0700:EN:NOT.

EC (2008). Progress report on the single European electronic communications


market 2007 (13th report). Retrieved from http://ec.europa.eu/information_society/policy/ecomm/doc/library/annualreports/13th/com_2008_153_en_fi nal.pdf.

EITO (2009). Data services give the wireless market momentum. Retrieved 21 March 2009 from http://www.eito.com/reposi/PressReleases/EITO- PR- 20090213.

FCC (2008). Auction of 700 MHz band licenses closes. Winning bidders announced for Auction 73. Retrieved from http://hraunfoss.fcc.gov/edocs_public/attach-match/DA- 08- 595A1.pdf.

Feijóo, C., J.L. Gómez- Barroso and A. Mochón (2009a). Reforms in spectrum management policy. In I. Lee (ed.), Handbook of Research in Telecommunications Planning and Management for Business (pp. 33–47). Hershey, PA: Information Science Reference.

Feijóo, C., I. Maghiros, F. Abadie and J.L. Gómez- Barroso (2009b). Exploring a heterogeneous and fragmented digital ecosystem: mobile content. Telematics and Informatics 3, 282–92.

Feijóo, C., I. Maghiros, M. Bacigalupo, F. Abadie, R. Compañó and C. Pascu (forthcoming). Content and Applications in the Mobile Platform: On the Verge of an Explosion. Seville: Institute for Prospective Technological Studies.

Forge, S., C. Blackman and E. Bohlin (2005). The demand for future mobile communications markets and services in Europe. Report No. EUR 21673 EN. Institute for Prospective Technological Studies – JRC – EC.

Forge, S., C. Blackman and E. Bohlin (2007). The mobile provide. Economic impacts of alternative uses of the digital dividend. Summary Report. SCF Associates.

Fransman, M. (2007). The New ICT Ecosystem. Implications for Europe. Edinburgh: Kokoro.

Giles, T., J. Markendahl, J. Zander, P. Zetterberg, P. Karlsson, G. Malmgren and J. Nilsson (2004). Cost drivers and deployment scenarios for future broadband wireless networks: key research problems and directions for research. Paper presented at the Vehicular Technology Conference, 2004. VTC 2004- Spring. 2004 IEEE 59th.

Gómez- Barroso, J. and A. Robles- Rovalo (2008). Wireless hopes for universal service in developing countries: an assessment of the Mexican context. info, 10(5–6), 83–91.

GSA (2009). 3G evolution steps confi rmed by multiple operators for HSPA+, LTE and EDGE evolution. GSM/3G network update. Available from http://www.gsacom.com/downloads/pdf/GSA_GSM_3G_Network_Update_Feb2009.php4.

Holden, W. (2008). Making music with mobile. Basingstoke: Juniper Research.Jaokar, A. and T. Fish (2006). Mobile web 2.0. London: Futuretext.Johansson, K., A. Furuskar, P. Karlsson and J. Zander (2004). Relation between

base station characteristics and cost structure in cellular systems. Paper pre-sented at the 5th IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC 2004.

Kocan, K.F., W.A. Montgomery, S.A. Siegel, R.J. Thornberry Jr. and G.J. Zenner (2002). Service creation for next generation networks. Bell Labs Technical Journal, 7(1), 63–79.

Modarressi, A.R. and S. Mohan (2000). Control and management in next- generation networks: challenges and opportunities. IEEE Communications Magazine, 38(10), 94–102.


Morris, A.C. (2005). Spectrum auctions: distortionary input tax or effi cient revenue instrument? Telecommunications Policy, 29(9–10), 687–709.

Nerandzic, D. (2008). Emerging technologies and their implications on regulatory policy. Paper presented at the 17th Biennial Conference of the International Telecommunications Society, The Changing Role of the Telecommunications Industry and the New Role for Regulation.

Netsize (2008). Netsize Guide. Mobile 2.0, You are in Control. Paris: Netsize.NGMN Alliance (2006). Next generation mobile networks beyond HSPA &

EVDO. Available from http://www.ngmn.org/uploads/media/White_Paper_NGMN_Beyond_HSPA_and_EVDO.pdf.

Ofcom (2008a). Progress on key spectrum initiatives. A review and update of the SFR and SFR:IP. Retrieved from http://www.ofcom.org.uk/radiocomms/sfr/sfrprogress/sfrprogress.pdf.

Ofcom (2008b). Trading guidance notes (December 2008). Retrieved from http://www.ofcom.org.uk/radiocomms/ifi /trading/tradingguide/tradingguide.pdf.

Oliver and Ohlbaum Associates (2008). The eff ects of a market- based approach to spectrum management of UHF and the impact on digital terrestrial broadcast-ing. Executive Summary.

Ouvrier, S. (2008). Driving key technologies for next generation mobile networks. Paper presented at the ICT Mobile Summit 2008.

Pascu, C. (2008). An empirical analysis of the creation, use and adoption of social computing applications. Report No. EUR 23415 EN – 2008. Seville: Institute of Prospective Technological Studies.

Ramos, S. (2005). Contribución al estudio, caracterización y desarrollo del sector europeo de comunicaciones móviles e Internet móvil. Unpublished PhD Dissertation, Universidad Politécnica de Madrid.

Ramos, S., C. Feijóo and J. Gómez- Barroso (2004). Barriers to widespread use of mobile Internet in Europe: an overview of the new regulatory framework market competition analysis. Journal of the Communications Networks, 3(3), 76–83.

Ramos, S., C. Feijóo and J. Gómez- Barroso (2009). Next generation mobile network deployment strategies. Journal of the Institute of Telecommunications Professionals, 3(1), 13–19.

Ramos, S., C. Feijóo, L. Castejón, J. Pérez and I. Segura (2002). Mobile Internet evolution models. Implications on European mobile operators. Journal of the Institution of British Telecommunications Engineers, 1(2), 171–6.

Reding, V. (2008). Digital Europe: the Internet mega- trends that will shape tomor-row’s Europe. Retrieved from http://europa.eu/rapid/pressReleasesAction.do?reference=SPEECH/08/616&format=HTML&aged=0&language=EN&guiLanguage=nl.

Ring, S. (2008). IMT- Advanced, a global platform for the next generations of mobile services. Paper presented at the International Converging Mobile Media Conference.

Sabat, H.K. (2002). The evolving mobile wireless value chain and market struc-ture. Telecommunications Policy, 26(9/10), 505–35.

Schwarz- da- Silva, J. (2008). Mobile research at the crossroads. Paper presented at the ICT Mobile Summit 2008.

317

15. Mobile Internet developments in Europe, East Asia and the US

Morten Falch, Anders Henten and Karsten Vandrup

15.1 INTRODUCTION

The chapter fi rstly examines how European countries are positioned with

respect to the diff usion of new mobile and wireless data technologies and

services as compared to the two most advanced East Asian countries,

Japan and South Korea, and to the US and Canada. Secondly, the chapter

discusses central factors which contribute to the explanation for the dif-

ferences in the development of the countries and regions in question. For

a period of time, Europe dominated the mobile scene with the second

generation Global System for Mobile communications (2G GSM) tech-

nology, which was developed as a European standard and spread to large

parts of the rest of the world. But with regard to the early versions of data

services on mobile platforms (for example, i- mode, Wireless Application

Protocol (WAP) and general packet radio service (GPRS) – often termed

2.5G), Japan and Korea took the lead, and the same applies to data serv-

ices on 3G platforms and beyond (3.5G). The US, on the other hand, has

trailed behind Japan and Korea and also the leading countries in Europe

regarding mobile technologies and services, but is clearly the world leader

in networked information technology (IT) on fi xed platforms. Lately,

however, the US seems to be catching up with Europe in the mobile fi eld.

These developments are what the chapter aims at discussing in terms of

facts and explanations.

Europe is a diverse market, including Scandinavia, being in the lead with

respect to the use of information and communication technology (ICT),

as well as countries in Eastern and Southern Europe, where the take- up

of ICT is generally less developed. Therefore, the European average is not

the best indicator to be used for the analysis of the European development.

We have, consequently, chosen to use data from four major European

economies instead: France, Germany, Italy and the UK.


Section 15.2 presents the conceptual framework of the chapter. This is

followed by section 15.3 comparing the status of mobile data in the three

regions. Section 15.4 analyzes the key segments of the mobile ecosystems.

Finally, conclusions are drawn in section 15.5.

15.2 BACKGROUND

At the general level, the conceptual framework of the chapter relates to

the concepts of systems of innovation (for example, Edquist, 2005) and

ecosystems (for example, Fransman, 2007). In fact, systems of innovation

approaches and ecosystem approaches are very similar, though systems of

innovation – as the concept denotes – gives emphasis to innovation, while

ecosystem approaches may have a broader scope. However, the focus

of ecosystem analyses is also innovations, as emphasized by Fransman

(2007). Moreover, the ecosystem concept connotes the right evolution-

ary setting for the development of technologies and services in a broader

environment. What the chapter aims at analyzing is the development of

specifi c types of ‘species’ (new mobile and wireless data technologies and

services) in diff erent ecosystems or environments.

In addition, diff usion theory (for example, Rogers, 2003) is used.

Diff usion processes are seen as a subset of the many processes and elements

analyzed in systems of innovation and ecosystem analyses. Diff usion anal-

yses obviously provide a deeper insight into diff usion processes as these

are the specifi c topics of diff usion analyses. While the literature on diff u-

sion traditionally is centered on examining factors determining demand,

the literature on systems of innovation and ecosystems often includes

supply as well as demand and a wide range of technological, organiza-

tional and institutional elements. Furthermore, the systems of innovation

and ecosystem approaches quite naturally point at comparative analyses

of diff erent systems (in diff erent regions, in our case).

At the more specifi c level, the purpose of the chapter is to examine the

development of mobile and wireless data technologies and services in a

selection of European countries (France, Germany, Italy and the UK),

Japan and South Korea, and the US and Canada, respectively, and to

discuss the possible central explanations for the diff erences in develop-

ments. What immediately springs to mind is the early and rapid develop-

ment of mobile Internet in Japan and Korea as compared to the European

countries as well as the US. Also, the strong position of Short Message

Service (SMS) in the European markets and a similar development in the

US market somewhat later is remarkable. Explanations for these facts

revolve around two phases in the developments: the fi rst phase where the

Mobile Internet developments in Europe, East Asia and the US 319

subsequent development is set ‘on track’; and the following phase where

‘the train is rolling’ and it becomes more diffi cult to change track. There is

path- dependency.

The fi rst phase has been debated and analyzed, to a relatively large

extent, with respect to comparisons of the developments in East Asia and

the developments in Europe and the US respectively (Funk, 2007, 2009;

Henten et al., 2004). Funk (2007) provides an explanation for the early

and swift development of mobile Internet in Japan, and soon thereafter

in Korea. Funk focuses on three main explanatory factors (for the success

of the Japanese i- mode system starting in 1999): (1) the implementation

of a micro- payments system; (2) a central role of entertainment services;

and (3) the deployment of an integrated approach including custom- made

end- user devices. Funk discards what he calls cultural explanations and

social factors relating to commuting and the relatively low diff usion of

fi xed Internet in Japan, and he gives emphasis to structural factors (verti-

cal integration) and business model elements regarding the services off ered

(entertainment) and the fi nancial design (micro- payments).

The present chapter also concentrates on structural explanations. While

the fi rst generation of mobile communications and the second generation

were based on a modular approach, which was an important reason for

the huge success of the European- based GSM standard, mobile Internet

requires a more integrated approach, in the initial phases at least (Funk,

2009). This argument is in line with the Henderson and Clark (1990) dif-

ferentiation between modular and architectural innovations. An architec-

tural (or systemic) innovation requires a more integrated approach in its

implementation phase but may later allow for modular approaches once

the innovation has gained ground and has settled in the market. While

Japan and Korea have followed an integrated approach, Europe and the

US have followed a more modular approach – in extension of their 1G and

2G models. Custom- made end- user devices, therefore, did not reach the

European and US markets in the early phases of mobile Internet, and the

WAP standard, which was implemented in the European and US markets,

became subject to several diff erent versions and never became the technol-

ogy platform envisioned in spite of the massive hype around it.

One of the explanations for these diff erences in development is related

to the relative strengths of the service providers (operators) and equip-

ment and device manufacturers in the Japanese and European markets

respectively. In Japan, the operators, and especially NTT, were relatively

large companies in comparison with the manufacturers and they could,

therefore, set the agenda and require specifi c systems and terminals. In

Europe, large trans- European and transnational manufacturers were one

of the results of the liberalization processes, and they opted for a strategy


where their equipment and devices could be sold all over Europe and the

world at large. The nationally confi ned telecom operators in the European

countries did not have the same say as their counterparts in Japan.

In the second phase, when there are a suffi ciently large number of users

of a system, other mechanisms become relevant. Funk (2007) deals with

these issues in his chapter entitled ‘Solving the startup problem in Western

mobile Internet markets’. Network eff ects (Economides, 1996; Varian and

Shapiro, 1999) including feedback mechanisms and installed bases, are a

central explanatory factor. In Japan and Korea, e- mail systems, which are

Internet compatible, became the text- based forms of mobile communica-

tion. In Europe, SMS became the dominant mode of text- based communi-

cation on the mobile platform, and SMS has later also gained considerable

ground in the US market as well. SMS has acquired such a huge following

in the European markets that an obvious hypothesis is that it is blocking

the development of Internet- based services on the mobile. This could thus

be considered as an occurrence of path- dependency (David, 2001) raising

barriers to the development of mobile Internet.

There are two aspects of this. One aspect has to do with the core of

network economics regarding direct network externalities or eff ects, that

is, that the more users there are of an interactive communication system,

the higher the utility for the individual user. If an e- mail- based system is

not compatible with the SMS system, users will be reluctant to change to

mobile e- mail. Furthermore, the users get habituated with and develop

competences in the specifi c types of systems they use – an argument in line

with David’s (2007) story concerning the QWERTY typewriter keyboard.

The other, and perhaps more important aspect, concerns the fact that

operators off ering SMS earn a great deal of money on the SMS system.

Not only are charges for sending normal SMS messages still high in many

cases in Europe, but the turnover and profi ts from premium services

are also high. The incentive to change to text communication based on

Internet is, consequently, low.

It should be mentioned that there are many advanced services off ered

via SMS systems. SMS is not as primitive a system as it is sometimes

portrayed. There are many diff erent development possibilities on top of

SMS systems. Nevertheless, the SMS- based forms of mobile data com-

munications constitute a side road as compared to Internet- based data

communications.

The question is where the US is positioned now with respect to these

phases and developments. There is no question that Japan and Korea are

still in the lead regarding mobile Internet. But how is the US positioned

when compared to Europe? This is what the chapter aims at shedding

light on, with available statistical data and explanations. The US never


implemented a single mobile standard as the Europeans did with GSM.

This seems to have held back the development in the US in the early

phases of the 2G developments and it has not, to any large extent, led to

integrated approaches between specifi c operators and specifi c manufactur-

ers, as in Japan and Korea. Recently, however, US- based companies like

Apple (iPhone) and Google have forcefully entered the mobile fi eld.

This indicates an important explanatory factor for the US catching up

with Europe – and eventually perhaps the East Asian countries. The US

has for many years had a lead in IT and Internet services on fi xed plat-

forms, especially on the supply side but also on the demand side. Many of

the world’s largest IT companies – whether in equipment or services – are

US- based. The issue is whether the US is able to leverage this advantage

onto the mobile and wireless fi eld.

This discussion is concerned with the overall structural level and the

possible explanations for the diff erences between Europe, East Asia and

the US. With respect to the diff usion subset, diff usions of innovations

follow an s- curve, according to Rogers (2003), where the early adopters

represent about 15 percent of the population. Early adopters will often be

characterized by being well educated and having a high income. This was,

for instance, the case with the adoption of the Internet in its early phase

(Rogers, 2001). Diff erences in income and education can also be used to

explain regional diff erences in the diff usion of ICT services. The countries

with the highest rankings in various e- readiness indexes are in general

countries with high levels of income and education. According to Rogers,

fi ve attributes of an innovation aff ect the speed of adoption (Rogers,

2003): relative advantage, compatibility, complexity, observability and

trialability.

Relative advantage is by far the most important of these factors and

explains about 80 percent of the diff usion. Rogers sees these as techni-

cal attributes, that is, as attributes that depend on the technical charac-

teristics of a specifi c innovation. However, the impact of each of these

attributes depends on the market environment in which the innovation is

being adopted. This is better refl ected in the technology assessment model

(TAM) introduced by Davis (1989). This model distinguishes between two

diff erent factors: perceived usefulness and perceived ease of use. These

relate to the same aspects as relative advantage and complexity but take

the user experience rather than the technology as the point of departure.

This is important especially in the present context, where we focus on

diff erences in the rates of adoption of the same or similar technologies in

diff erent regions.

This model has been applied in a study of user acceptance of advanced

mobile services in the Dutch market (López- Nicolas et al., 2008). The


study confi rms that the TAM model is useful for explaining service adop-

tion. In addition, the study documents the importance of social infl uence

on perceived usefulness as well as perceived ease of use.

The decomposed theory of planned behavior (DTPB) builds on the

same two parameters but adds Rogers’s compatibility as a third parameter

(Hernandez and Mazzon, 2007). The relative advantage or the perceived

usefulness of mobile data access depends on what kinds of services they

enable, and what the alternatives are. Furthermore, the concept of com-

patibility relates not only to compatibility of technical standards, but also

to compatibility with existing work practices.

The issue of pricing is not really taken into account in these models. But

pricing is an important issue when it comes to the adoption of new services

on the consumer market. This is addressed in the value- based adoption

model (Kim et al., 2007), which is developed especially for analyzing adop-

tion of mobile Internet services. This model operates with four diff erent

factors aff ecting the perceived value and the adoption intention: useful-

ness, enjoyment, technicality and the perceived fee. It is shown that all four

factors have a statistically signifi cant impact.

15.3 STATUS FOR THE DIFFUSION OF MOBILE DATA

The Nordic countries were the fi rst to achieve a wide penetration of mobile

services. However, it was not until the introduction of GSM as a common

European standard that Europe became a leader in mobile communica-

tion. Up to 1998, the penetration of mobile phones was lower within the

EU than in the US, but in 1999 the EU penetration increased to 39.6 per

100 inhabitants compared to 31.5 in the US. In 2000, the EU penetration

jumped to 63.4. Thereby, the penetration became higher than in Korea

(57) and Japan (52.6) – the leading countries in the Asian markets (OECD,

2003).

The successful implementation of one common standard for 2G teleph-

ony throughout Europe created a mass market for terminals and enabled

roaming in most parts of Europe. Europe dominated the top ten list

regarding the penetration of mobile phones, and GSM became the domi-

nant world standard for 2G.

As shown in Table 15.1, the penetration in Europe has remained higher

than in the US and in East Asia. However, regarding further mobile gen-

erations, the European dominance has dwindled. While NTT DoCoMo

was successful in its implementation of i- mode at the turn of the century,

European countries were far from successful in their introduction of data


services based on the WAP standard at the same time. South Korea has

been another early adopter of mobile data services, which now in many

ways are more advanced than in Europe. Already in 2005, an overwhelm-

ing majority of the subscribers in South Korea had switched to 3G, and

in Japan, 50 percent of the subscribers had done the same (OCED, 2007).

At that time, most European countries had not yet introduced 3G at all

and it was only in Italy and Austria that the penetration of 3G was more

than 3 per 100 inhabitants. Although the penetration of 3G is presently

increasing rapidly in Europe, the development is still far behind Japan and

Korea.

The development in the penetration of 3G in the USA is comparable

to that in many European countries. It should, however, be noted that

although the percentage of 3G subscribers is comparable or even higher

than in the European countries, the total penetration of mobile subscrib-

ers is still lower. The development in Canada lacks behind any of the other

countries included in Table 15.1.

Although related, 3G and data services are two diff erent things. 3G

services include voice as well as data. In fact, voice ARPU (average

revenue per user) is larger than data ARPU for 3G subscribers also, and

from an operator’s point of view an important function of 3G networks

is that they can expand the capacity for mobile telephony in high- density

areas. Moreover, some data services, for example, SMS, can run on both

2G and 3G networks.

Table 15.1 Mobile subscribers per 100 inhabitants in selected countries

1996 1998 2001 2003 2005 2008

3Q

3G

subscriber

(%)

USA 16.6 25.6 45.1 54.5 71.8 87.6 22.8

Canada 11.5 17.7 34.9 41.8 51.64 62.6 9.0

France 4.2 19.2 62.6 67.7 76.71 91.0 18.5

Germany 7.1 17 68.3 78.5 96.04 122.0 19.2

Italy 11.3 35.6 87.1 97.6 122.16 149.3 32.2

UK 11.6 21.9 77.1 89.2 106.26 122.4 25.9

Japan 21.4 37.4 58.8 67.9 75.51 82.3 81.3

South

Korea

7.0 30.2 61.4 70.1 79.39 92.6 72.0

Sources: OECD (various years) and Netsize (2009).


Mobile data is a relatively new service area and it is, therefore, diffi cult

to fi nd international comparable data in the offi cial statistics. The most

comprehensive source publicly available is the Netsize Guide 2009 (Netsize,

2009), which includes country information on mobile data services. Table

15.2 is constructed on basis of data from various tables in that report.

According to Netsize (2009), North America is in absolute terms the

region that generates the highest income per subscriber on mobile data

(Figure 15.1). This somewhat surprising information is supported by other

reports, for instance Fierce Mobile Content (2007). However, before the

US is nominated as the leader in mobile data, a few reservations have to

be expressed. First of all, Asia Pacifi c includes other countries than Japan

and Korea. Both countries have a higher ARPU than the US (Japan has

a much higher ARPU). Also in the UK the ARPU is higher than in the

US. Secondly, as noted above, penetration is lower in the US and Canada

than in Europe. Higher penetration can lead to a lower ARPU as more

users with a limited service demand will be included. In addition, high

penetration rates refl ect that more users have more than one subscription.

This is obviously the case in Italy, for example, where the penetration is

almost 150 percent. According to a study by Strategy Analytics (2008),

68 percent of the population in North America had at least one mobile

phone in 2006, while the same fi gure for Western Europe was 80 percent.

This indicates that double subscriptions are of little importance in North

Table 15.2 Penetration of mobile subscribers and average revenue per

user (ARPU) (euros)

Pene-

tration

rate, 3Q

ARPU ARPU

Mobile

messaging

ARPU

Mobile content

ARPU

Mobile data

absolute % absolute % absolute %

USA 87.6 35.47 3.38 9.5 4.33 12.2 7.71 21.7

Canada 62.6 41.76 3.29 7.9 3.23 7.7 6.52 15.6

France 91.0 36.13 3.91 10.8 2.72 7.5 6.63 18.4

Germany 122.0 17.15 3.63 21.2 2.07 12.1 5.7 33.2

Italy 149.3 21.42 5 23.3 2.75 12.8 7.75 36.2

UK 122.4 29.87 8.83 29.6 2.8 9.4 11.63 38.9

Japan 82.3 34.81 6.97 20.0 7.5 21.5 14.47 41.6

Korea 92.6 26.46 2.96 11.2 4.62 17.5 7.58 28.6

Source: Netsize (2009).


America, while phone number two or three may account for as much

as 30 percent of the subscriptions in Western Europe. Calculations of

income per inhabitant may decrease or even annihilate the gap between

the US and Canada and Western Europe. Finally, ARPU depends on

pricing as well as use.

It is, nevertheless, interesting to note that not only the level but also the

growth of ARPU is higher in North America than in the other regions.

For instance, the mobile data traffi c in T- Mobile’s US network more than

doubled in one year from 1Q 2007 to 1Q 2008 leading to a 31 percent

growth in data revenue (Fitchard, 2009).

Another interesting observation, which might be related to this, is the

distribution of revenues generated from messaging and content services.

The US is the only country where content services dominate. Mobile mes-

saging includes SMS services and the use of SMS had a slower start in the

US than in Europe. According to Strategy Analytics (2008), 50 percent

of the mobile users in North America only used voice in 2006. This is

compared to 10–20 percent in Western Europe and Asia Pacifi c. This dif-

ference refl ects a much more sparse use of SMS in North America at that

time. But high growth rates in recent years have compensated for this, so

the use per subscriber was at the beginning of 2009 almost double that in

Europe (Mobile Europe, 2009).

10

9

8

7

6

5

4

3

2

1

0

USA/Canada Western

Europe

Eastern

Europe

Asia Pacific

20073q 2008


Figure 15.1 Data ARPU(US$ per month)


Although the number of SMS messages continues to grow, its share in

non- voice revenues is declining. According to Netsize (2009), SMS used to

generate up to 80 percent of the non- voice revenues for a European opera-

tor in the early 2000s, but this had by 2008 declined to about 50 percent at

least for some of the operators.

When it comes to the number of users of other mobile services than

SMS, the US and Europe are far behind the most advanced Asian coun-

tries. Capgemini (2008) reports that in 2006 and 2007, 89 percent of the

Japanese subscribers used mobile Internet services at least once a month.

In the US, only 12 percent were using such services. In Europe, this fi gure

increased from 11 percent in 2006 to 14 percent in 2007.

According to Nielsen Mobile (2008), 37 percent of the mobile subscrib-

ers in the US paid a subscription fee for mobile Internet by the fi rst quarter

of 2008. However, only 15.6 percent are active users. This compares to

slightly lower fi gures for the UK (12.9 percent) and Italy (11.9 percent),

and somewhat lower fi gures in France (9.6 percent) and in Germany (7.4

percent) (Table 15.3). The lead of the US compared to Europe is due to

a growth in the number of users of more than 100 percent from 2007 to

2008. This is compared to a growth in Europe at 50 percent (highest in

France with 69 percent) (IT Facts, 2009).

15.4 COMPARISON OF MOBILE ECOSYSTEMS

The central market actors in the mobile ecosystem are the following:

equipment and device manufacturers, software developers, network oper-

ators, service providers, producers and providers of content, and users.

Table 15.3 Mobile Internet usage across 16 measured countries

(Q1 2008) (%)

Country % Country %

US 15.6 China 6.8

UK 12.9 Philippines 3.4

Italy 11.9 Singapore 3.0

Russia 11.2 Brazil 2.6

Spain 10.8 Taiwan 1.9

Thailand 10.0 India 1.8

France 9.6 New Zealand 1.6

Germany 7.4 Indonesia 1.2

Source: Nielsen Mobile (2008).


In this section, we will limit our analysis to the following segments, as we

see these as being the most important in this context: mobile terminals;

network infrastructures; services and content; and users.

15.4.1 Mobile Terminals

Ease of use is an important parameter for the adoption of a new service.

If the majority of handsets in the market have none or only hard- to- use

Internet browsing functionalities, Internet browsing will be low. Until

2004, only a few smartphone handsets with a user- friendly browsing func-

tionality were sold in the US, whilst more models were available in the

European markets. Together with a late rollout of US- wide 3G networks

and data services this meant that the US was lagging behind Japan, Korea

and Europe.

It follows from Table 15.4 that the European and the US markets are

very diff erent with regard to mobile terminal preferences. While Nokia

dominates the European market, American manufacturers dominate

the US market. In spite of Nokia’s global success, it has never been able

to penetrate the US market as heavily as it has done in the rest of the

world. While Nokia’s market share worldwide equaled 40 percent in 2008

(Gartner, 2009), its share in the US was only 10 percent for the same year

Table 15.4 Top devices: mobile Internet users

Top devices: % of mobile Internet users (US), Q1 2008

Device %

1 Motorola RAZR/RAZR2 10

2 Apple iPhone 4

3 RIM BlackBerry 8100 series (pearl) 2

4 RIM BlackBerry 8800 series (8820, 8830) 2

5 Motorola Q Series (Moto Q, 9h, 9c, 9m, Q Glo) 2

Top devices: % of mobile Internet users (EU), Q1 2008

Device %

1 Nokia N95 5

2 Nokia N70 4

3 Motorola RAZR/RAZR2 3

4 Sony Ericsson K800i 3

5 Nokia N73 3

Source: Nielsen Mobile (2008).


(Digital Trends, 2009). Nokia has been engaged in a long- term patent

dispute with the rival mobile technology company Qualcomm over licens-

ing of code division multiple access (CDMA) technology, and some indus-

try observers have speculated that the legal wrangling has made Nokia

reluctant to commit to the still CDMA- dominated US market (Helsinki

Sanomat, 2008).

With the deployment of RIM’s Blackberry starting in 2000, and later

with Apple’s introduction of the iPhone in 2007, the American market

got its own drivers for mobile broadband services. The Blackberry was a

simple way of getting e- mails on a handset. This was made available fi rst

by using standard GSM data rates (max. 13Kbps) and later using 2.5G

services like GPRS (70Kbps) and EGPRS (enhanced GPRS) (473 Kbps)

data enhancements. The Blackberry fast became the icon of the American

business people, allowing them to stay updated on the mailbox while

traveling. RIM claimed to have 1 million subscribers in 2003 and 2 million

in 2004, increasing to 21 million in 2008 (Impact Lab, 2009). Table 15.4

makes it clear that the top- selling smartphones in the US are the American

brands such as Motorola, RIM and Apple, whereas the European market

favors European brands. Another reason why the mobile e- mail service

was a blockbuster in the US in particular was the lack of SMS use in the

US at that point of time. In Europe, text messaging via SMS was and still

is by far the most- used non- voice service, and also the service that brings

most revenue to the operators.

Blackberry Connect from RIM, Apple’s proprietary e- mail platform,

or off erings coming with the Symbian and Windows Mobile OS, are the

most widespread platforms for mobile e- mail. The most popular devices

are those where mobile data functionality has been given high priority in

the design. Such phones can be expected to generate far more data traffi c

than general- purpose devices.

According to Capgemini (2008), only 13 percent of the mobile phone

users accessed the mobile Internet in January 2008. However 58 percent

and of the mobile users equipped with a smartphone and 85 percent of

iPhone users reported that they accessed the mobile Internet within the

same period. In particular, the introduction of the iPhone has had a posi-

tive impact on the usage of mobile Internet services. Hansberry estimates

that as much as two- thirds of the mobile Internet traffi c is generated by use

of iPhones (Hansberry, 2009).

It should be noted that mobile phones are far from the only termi-

nals to be used for mobile data services. In fact one of the key drivers

for the growth in mobile data services within the past year has been

use of Universal Serial Bus (USB) modems connecting laptops to the

Internet.


15.4.2 Network Infrastructures

Mobile data can be provided by the use of a wide range of diff erent

network technologies. Although each standard has its own charac-

teristics, the network technologies are often grouped into generation:

2G (for example, CDMA and GSM), 2.5G (for example, CDMA2000

1RTT – Radio Transmission Technology – and GPRS), 3G (for example,

CDMA2000 Evolution- Data Optimized – EVDO; and wideband CDMA

WCDMA), and 3.5G (for example, High Speed Packet Access HSPA).

Simple data services such as SMS can run on 2G (13 Kbps1), but other

data services demand at least 2.5G (70–473 Kbps).

Second generation was in the beginning a larger success in Europe com-

pared to the US. Where Europe agreed to deploy the same technology –

the 900 MHz GSM system – diff erent standards were used in the US. The

outcome was that no single handset could be used everywhere in the US.

This stalled the deployment of 2G handsets in the US and also resulted in

a later uptake of the following 3G systems.

A likely reason for the slower uptake in the US of 3G technology is that

the US has been in the forefront in the deployment of the IEEE 802.11

wireless local area network (LAN) standards (Wi- Fi) and that users in

this way have had data services available ‘out- of- offi ce’ to some extent.

Obviously, Wi- Fi is not a mobile technology. However, as it is provided

widely in some areas and regions of the US, access to the Internet, for

example, from a Wi- Fi based personal digital assistant (PDA) or laptop,

is possible in many places, holding back a direct demand for a cellular

solution.

Japan was the fi rst to introduce 3G in 2001, and also South Korea was

an early adopter (2003). In the major European countries, 3G was intro-

duced in 2003–04, while the US and Canada came a bit later (AT&T as

the fi rst US operator introduced 3G in 2004). When it comes to 3.5G, the

diff erences are much smaller. AT&T, for instance, introduced High- Speed

Downlink Packet Access (HSDPA) in December 2005, half a year in

advance of most European countries (www.3G- americas.org, 2009).

Looking a few years ahead, two technologies are expected to domi-

nate: Long- Term Evolution (LTE) and Worldwide Interoperability for

Microwave Access (WiMAX). The expectation has been that WiMAX

will dominate in the American and some Asian markets, while LTE will

dominate in Europe and in Japan (Vandrup, 2008). However, at least

two American operators (Verizon and AT&T) have announced that they

will go for LTE. Verizon signed an LTE network deal with Ericsson in

February 2009. The American market will thus have two competing tech-

nologies, as will a number of other markets.


15.4.3 Services and Content

As the number of mobile users has saturated in the economically devel-

oped countries and as the ARPU for voice has decreased, operators have

been seeking new revenue streams off ering mobile data services. SMS has

been a lucrative revenue generator especially in Europe but other non-

voice services are also increasingly growing.

The highest data ARPU is found in the US, but the Asia Pacifi c coun-

tries have the highest share of their total mobile turnover deriving from

mobile data. The global average is around 20 percent according to The

Netsize Guide (Netsize, 2009). The Asia Pacifi c average is almost 25

percent while the Western European and the US and Canada averages are

about 21–22 percent.

However, the share of mobile data revenues as compared to total mobile

revenues is increasing faster in the US than in the Asia Pacifi c countries

or the European countries. From the fi rst quarter of 2007 to the second

quarter of 2008, mobile data revenues increased from 15 percent to 21

percent in the US, while they grew from 18.5 percent to 21.5 percent in

Europe and from 23 percent to 24.5 percent in Asia Pacifi c (Netsize, 2009).

This equals an increase of 40 percent in the US – Canada, 16 percent in

Western Europe, and 6.5 percent in the Asia Pacifi c.

As to the categories of mobile data services, Table 15.5 presents the

distribution of diff erent kinds of non- voice services for a group of North

American, Western European and East Asian countries.

Table 15.5 shows that Internet access is by far the most important

mobile data service. Internet access includes many diff erent services but

Table 15.5 Distribution of revenues in mobile content services, 2008 (%)

Music Games TV Video Internet Other Total

US 5.8 3.4 7.0 2.0 72.4 9.4 100.0

Canada 7.1 5.9 0.7 1.3 75.0 10.0 100.0

France 11.2 5.9 1.2 2.5 66.7 12.3 100.0

Germany 9.9 5.3 1.5 2.0 57.7 23.8 100.0

Italy 9.7 6.9 3.6 2.8 56.1 20.9 100.0

UK 7.9 7.1 0.8 3.1 56.2 24.8 100.0

Japan 4.0 3.4 2.3 2.3 75.7 12.3 100.0

South Korea 17.3 7.6 7.4 4.7 40.6 22.4 100.0



denotes general best eff ort (quality of service) Internet access. The other

service areas are dedicated services managed by the mobile operators. The

most surprising observation is possibly the great similarity between all the

countries listed with respect to the distribution between the mobile data

service categories. Internet access is, however, at a higher level in North

America and Japan as compared to Western Europe. Moreover, South

Korea presents a slightly diff erent picture because of the great weight of

music. Also, the development of mobile TV varies among the countries

listed, refl ecting the number of mobile TV users in the diff erent countries.

The exception is Japan, which is one of the most advanced countries with

respect to mobile TV services. The reason may be that the table measures

revenues and not usage, and that revenues from free- to- air mobile broad-

cast services are limited.

The comparison of usage data from Europe with similar data from the

US is also illustrative (Table 15.6). From this table it follows that the US is

leading in Internet- based services such as mobile e- mail, and that ‘info via

browser’ is more widely used in the US, while received SMS ads and music

are more widespread in Europe. This could indicate that Europe, being an

Table 15.6 Internet activity of European and US mobile subscribers in

November 2008 (%)

US* Germany France Italy UK

Sent/received photos

or videos

28 23 25 30 30

Received SMS ads 24 27 64 54 36

Accessed news/info

via browser

18 9 14 10 19

Used e- mail 17 9 9 12 12

Listened to music 9 20 17 17 22

Accessed social

networks

9 3 9 5 9

Played downloaded

game

9 8 4 8 10

Purchased ringtones 9 4 3 4 3

Accessed downloaded

application

8 4 3 6 6

Watched video 3 5 6 7 4

Note: *September 2008

Sources: Mobile penetration: how many people access a mobile site at least once a week – itfacts.biz, 29 September 2008; and Internet activity of European mobile subscribers in November 2008 – itfacts.biz, 23 February 2009.


early adopter of mobile data services via SMS, has had more emphasis on

developing applications which can be provided without Internet access. In

the US, where SMS was adopted later, while e- mail and web access were

introduced earlier on, the market for such SMS applications has been

more limited – especially in the business market.

One of the main discussions concerning the diff erences in the develop-

ment and take- up of mobile data services in diff erent regions of the world

has dealt with business models. DoCoMo in Japan, for instance, had an

early success with the semi- walled i- mode business model. Table 15.5,

however, indicates that there are no consistent diff erences in the openness

of business models among the countries listed. In all countries, general

Internet access makes up the largest portion and the managed services are

generally at a comparable level among countries.

15.4.4 Adoption of Mobile Data Services by Users

The rate of adoption depends on the perceived usefulness of mobile data

services compared to the expected costs related to acquiring, learning to

use the service, and compatibility with current practice. The importance of

these parameters has been confi rmed by the empirical studies on the adop-

tion of mobile data services mentioned above (López- Nicolas et al., 2008;

Kim et al., 2007).

The perceived usefulness depends on the services available and possible

substitutes. With regard to messaging services, SMS is an obvious sub-

stitute to e- mail via mobile Internet access. Thus the availability of SMS

reduces the usefulness of mobile Internet access.

SMS and to a certain extent mobile e- mail are substitutes to mobile

voice telephony. A major driver for the adoption of SMS has been a com-

parative cost advantage. The cost advantage of e- mail and SMS services

depends on the pricing schemes applied. With a fl at rate on mobile data

services, such services are attractive once a subscription is made.

It is a subject for discussion whether fi xed Internet services complement

or substitute for mobile data services. A clear analogy is the relationship

between fi xed and mobile telephony. In its early phases, mobile telephony

stimulated use of fi xed phones as well. It became possible to call people

from fi xed phones even if the recipients were on the move. In a later phase,

the mobile phone became the primary means of communication, and this

has resulted in an ongoing reduction in the number of fi xed telephone

network subscribers.

When it comes to access to fi xed network services, the number of per-

sonal computers (PCs) and Internet users is highest in North America and

lowest in Europe. With regard to broadband penetration, the picture is


more mixed. The highest penetration is to be found in South Korea and

the lowest in Italy and Germany (Figure 15.2).

A high penetration in the use of fi xed network services can have a

positive impact on the perceived usefulness of mobile data services as well.

Mobile Internet access enables access to the fi xed services, and a developed

market of such services adds to the usefulness of mobile access. In addition

service providers’ experiences can be used to develop services especially

designed for mobile terminals.

An often- used method to compare national markets for ICT service

is the construction of aggregate indexes for e- readiness. These indexes

are composed by a number of diff erent parameters on usage, prices and

business environment. The indexes indicate that North America, and

particularly the US, is better prepared for the implementation of ICT

services than the major European countries. In particular Italy, one of the

advanced countries with regard to many kinds of mobile services, ranks

low in international comparisons (Table 15.7).

An interesting point to be made in relation to an analysis of the more

disaggregated parameters on usage by individuals, business and govern-

ment is that countries with similar total rankings may have very diff er-

ent profi les (Table 15.8). Germany and Japan rank high in business use,

100806040200

South Korea

Japan

UK

Italy

Germany

France

Canada

USA

Broadband subscribers

Internet users

PCs

Source: World Economic Forum (2008).

Figure 15.2 PCs, Internet users, and broadband subscribers per 100

inhabitants


while all European countries except France rank low in government

use.

Many mobile data services address mainly the private consumer market

and, therefore, the individual usage ranking is the most relevant one. But

business and especially government usage are relevant to the evaluation of

the supply side, as the development of e- business and e- government serv-

ices depend on advanced usage in these sectors.

Table 15.8 E- readiness ranking in usage parameters

Individual Business Government Total

USA 17 8 5 9

Canada 12 16 16 15

France 23 18 10 19

Germany 21 2 38 22

Italy 25 45 47 33

UK 6 11 22 12

Japan 22 3 31 21

South Korea 4 15 7 3

Source: World Economic Forum (2008).

Table 15.7 E- readiness rankings for selected countries

E- readiness1 Ranking E- readiness2 Ranking

USA 8.95 1 5.49 4

Canada 8.49 12 5.3 9

France 7.92 22 5.11 21

Germany 8.39 14 5.19 16

Italy 7.55 25 4.21 42

UK 8.68 8 5.3 12

Japan 8.08 18 5.14 19

South Korea 8.34 15 5.43 13

Sources: 1. Economist Intelligence Unit (2008). 2. World Economic Forum (2008).


15.5 DISCUSSION AND CONCLUSION

Standardization has been a key to success in early mobile markets.

Scandinavia became the fi rst mass market for mobile telephony on the

basis of the common Nordic NMT (Nordic Mobile Telephone) standard.

Later, the successful promotion of GSM as a common European 2G

standard was a major reason for the European lead in mobile communica-

tion in the 1990s. Common standards enabled national and international

roaming and it facilitated an open and competitive environment leading to

low charges and high levels of use.

However, this open, standardized and modularized approach turned

out to be a problem for the development of mobile Internet. The fi rst suc-

cessful mobile Internet system was i- mode, which was implemented in a

vertically integrated system by NTT DoCoMo as the dominating actor in

Japan. Japan and Korea became the most advanced nations with regard

to the diff usion and use of mobile technologies. These two countries have

maintained this lead in the 3G–3.5G era, but the analysis provided in

sections 15.3 and 15.4 indicates that this lead has become less obvious in

certain areas.

The explications for the East Asian lead have been many. Funk (2007)

believes that analyses by European and US observers have had an errone-

ous focus on cultural factors and long commuting times and low fi xed

Internet penetration. Instead, Funk emphasizes the interplay between

technology and business model developments and builds on the theory

regarding advantages of integrated business models in the fi rst phases of

architectural or systemic innovations (Henderson and Clark, 1990).

The present chapter supports the emphasis on technology and business

models as the prime explanatory factors. What the chapter aims at doing is

to insert the technology and business model factors into the setting of the

diff erences in innovation systems or ecosystems in the regions in question.

As stated by Funk (2007, 2009), the Europeans as well as the Americans

have built on the successful 2G approach, but the reason for this (in the

European case, at least) lies not only in a mental model building on past

experience instead of analyses of the present and future development pos-

sibilities, but also on structural factors in the telecommunications sectoral

innovation systems and ecosystems in Europe with strong international

equipment and device manufacturers and, to a large extent, nationally

confi ned markets for operation. This provides a relation of strength

between manufacturers and operators, but it also sets the necessity of a

general standardized approach in order to overcome the limited size of the

national markets in Europe.

This discussion is concerned with the ‘stage- setting’ fi rst phase of


mobile Internet developments. Once the path has been laid out, factors

regarding network eff ects come into play in a second phase. Instead of

the Internet- compatible e- mail development for text- based messaging in

the Japanese market, the European markets saw a strong development

of SMS. The initial success of SMS bred more success for SMS, based on

network eff ects, and has settled into a system with widespread SMS com-

petences among the users and an interest in maintaining the SMS system

from the supply side because of the vast turnover and income from basic

and premium SMS. The result is a total dominance of the SMS system

for mobile data in Europe. And even though there are, indeed, advanced

off erings on top of the SMS system with premium services of many dif-

ferent kinds, and even though history shows that a dominant system may

quite quickly be toppled by other somehow superior systems (for instance

Word substituting for WordPerfect), the outcome is that Europe, at the

moment, is locked into a side road instead of the main Internet road for

mobile data developments.

In the US market, the 2G take- off was hampered by competing stand-

ards complicating, for instance, interstate roaming. And with respect

to 3G, the US still trails behind not only Japan and Korea but also the

more advanced countries in Europe. However, when it comes to the use

of mobile data services, the situation is less clear. In fact, the US has in

absolute terms the highest ARPU for mobile data services. As explained

in the chapter, a number of reservations must be expressed before it can

fi rmly be concluded on this background that the US has become the world

champion in this ‘discipline’. ARPU depends on use as well as pricing and

the number of subscribers, and in Europe, ARPU is lowered by the fact

that many customers have more than one subscription.

Nevertheless, the various parameters compared in the previous sections

indicate an ongoing convergence between conditions on regional markets.

While East Asia was well ahead with regard to the introduction of 3G

networks, the upgrade to 3.5G has taken place almost simultaneously in

the diff erent regions. When it comes to the implementation of WiMAX

and LTE, the US may take off ahead of Europe. Thus, a comparison of

the various parameters does not provide a clear picture of leaders and lag-

gards with respect to the adoption of mobile data services. However, it is

clear that the European lead in the mobile sector has been lost, fi rst to East

Asia and now, to a certain extent, to the US as well.

A key issue triggered by this chapter is, consequently, whether this is

due to an ongoing convergence based on the globalization in all parts of

the value networks, or whether there are some specifi c factors favoring

market development in the US as compared to market developments in

Europe.


A striking diff erence, just a few years ago, between the US market and

the markets in Europe was the relatively low use of SMS in the US. This,

however, does not apply any more. SMS use has grown extensively in

the US market (IT Facts, 2008). But in comparison with the European

markets, both ‘tracks’ are used in the US. E- mail via mobile Internet is

taken up primarily by business people, while SMS is used by the general

mobile users. The more extensive diff usion of Internet- based messaging

may promote other mobile Internet services as well. Once subscribers have

access to the mobile Internet, they begin to use some of the other services

(for example, content services), which are available.

Another and probably more important factor is the convergence

between the mobile and Internet sectoral systems of innovation. The devel-

opment and use of Internet services are more advanced in North America

than in the major European countries. This is refl ected in a better ranking

in various indexes on e- readiness, for instance the e- readiness index by the

Economist Intelligence Unit (EIU, 2009). It is likely that the US market

for mobile data leverages on the US lead in fi xed Internet services. The US

IT giants, Microsoft, Google and Apple, are forcefully entering the mobile

fi eld. While mobile Internet was driven by the mobile operators in Japan,

it may be the IT companies in the US which will be the drivers of mobile

Internet in the US.

In a survey performed by Netsize (2009), respondents could choose

between Internet giants, mobile operators, mobile pure players, plat-

form providers and handset vendors as the actors to dominate the

‘mobile 2.0’ development. Sixty- four percent and 46 percent respectively

backed Internet giants and mobile operators as the likely winners, while

the other players were at a considerably lower level (Netsize, 2009).

Although this only refl ects the opinion of industrial observers, it does

point at a trend towards the growing importance of Internet players

in the fi eld of mobile communications. With the US dominance in this

fi eld, it may be an indication of the future strength of the US in mobile

data developments.

The above- mentioned development indicates that Europe cannot take

a leading future position in mobile communications for granted. With

the development of mobile data services, the mobile system of innova-

tion becomes more interlinked with other parts of the ICT sector, and

it becomes more diffi cult to maintain a leading position in mobile com-

munications without a similar position within ICT more broadly speak-

ing. Policies addressing improvement of e- readiness in general might,

therefore, be an effi cient way to ensure continued development and use

of mobile services in Europe. Policies to promote mobile data need to

encompass a wide variety of policy areas as the development of mobile


data depends on structural factors in the markets on the supply side as

well as take- up factors like the general e- readiness of the potential users

on the demand side.

NOTE

1. Data rate fi gures are theoretical maximums, and in practice these fi gures are lower.

REFERENCES

Capgemini, (2008). Mobile Internet services in Europe and USA: initiatives to drive adoption and usage. Telecom and Media Insights, 30, 1–12.

David, P. (2001). Path dependence, its critics and the quest for ‘historical econom-ics’. In P. Garrouste and S. Ioannides (eds), Evolution and Path Dependence in Economic Ideas: Past And Present (pp. 15–40). Cheltenham, UK and Northampton, MA, USA: Edward Elgar.

David, P. (2007). Clio and the economics of QWERTY. American Economic Review, 75(2), 332–7.

Davis, F.D. (1989). Perceived usefulness, perceived ease of use, and user accept-ance of information technology. MIS Quarterly, September, 319–40.

Digital Trends (2008). Nokia wants bigger share of US market. Available at http://www.digitaltrends.com.

Economides, N. (1996). The economics of networks. International Journal of Industrial Organization, 14(6) 673–99.

Economist Intelligence Unit (EIU) (2008). E- readiness rankings 2008 – maintain-ing momentum. EIU, London.

Economist Intelligence Unit (EIU) (2009). E- readiness rankings 2009 – the usage imperative. EIU, London.

Edquist, C. (2005). Systems of innovation – perspectives and challenges. In J. Fagerberg, D. Mowery and R. Nelson (eds), Oxford Handbook of Innovation (pp. 181–208). Oxford: Oxford University Press.

Fierce Mobile Content (2007). North America now leads on data ARPU. http://www.fi ercemobilecontent.com. 22 October 2007.

Fitchard, K. (2009). T- Mobile data revenue growth due mostly to SMS, not 3G. Telephony Online. http://telephonyonline.com/wireless. 27 February.

Fransman, M. (2007). The New ICT Ecosystem: Implications for Europe. Edinburgh: Kokoro.

Funk, J.L. (2007). Solving the startup problem in Western mobile Internet markets. Telecommunications Policy, 31, 14–30.

Funk, J.L. (2009). The co- evolution of technology and methods of standard setting: the case of the mobile phone industry. Journal of Evolutionary Economics, 19, 73–93.

Gartner (2009). Worldwide mobile phone sales. Gartner, Egham. http://www.gartner.com. 2 March 2009.

Hansberry, E. (2009). iPhone claims nearly two- thirds of mobile browser share. http://www.informationweek.com. 3 March.


Helsinki Sanomat (2008). Nokian vaikeudet USA:ssa pahenivat. www.nettisano-mat.com. 11 November.

Henderson, R.M. and K.B. Clark (1990). Architectural innovation: the recon-fi guration of existing product technologies and the failure of established fi rms. Administrative Science Quarterly, 35(1), 9–30.

Henten, A., H. Olesen, D. Saugstrup and S.E. Tan (2004). Mobile communica-tions: Europe, Japan and South Korea in a comparative perspective. Info – The Journal of Policy, Regulation and Strategy for Telecommunications, 6(3), 197–207.

Hernandez, J.M.C. and J.A. Mazzon (2007). Adoption of Internet banking: proposition and implementation of an integrated methodology approach. International Journal of Bank Marketing, 25(2), 72–88.

Impact Lab (2009). The Blackberry story. Available at http://www.impactlab.com/2009/02/08/the- blackberry- story.

IT Facts (2008). Mobile penetration: how many people access a mobile site at least once a week. http://www.itfacts.biz. 29 September 2008.

IT Facts (2009). Internet activity of European mobile subscribers in November 2008. http://www.itfacts.biz. 23 February 2009.

Kim, H., H.C. Chan and S. Gupta (2007). Value- based adoption of mobile inter-net: an empirical investigation. Decision Support Systems, 43, 111–26.

López- Nicolas, C., F.J. Molina- Castillo and H. Bouwman (2008). An assessment of advanced mobile services acceptance: contributions from TAM and diff usion theory models. Information and Management, 45, 359–64.

Mobile Europe (2009). Mobile messaging in 2009 – SMS is still the driver. 5 February 2009. http://www.mobileeurope.co.uk/.

Netsize (2009). The Netsize Guide 2009. http://www.netsize.com.Nielsen Mobile (2008). Critical Mass – The Worldwide State of the Mobile Web.

http://mmaglobal.com.OECD (various years). Communications Outlook. Paris: OECD.Rogers, E.M. (2001). The digital divide. Convergence: The International Journal of

Research into New Media Technologies, 7(4), 96–111.Rogers, E.M. (2003). Diff usion of Innovations. New York: Free Press.Strategy Analytics (2008). Understanding the mobile ecosystem. http://www.

adobe.com/devnet/devices/articles/mobile_ecosystem.pdf.Vandrup, K. (2008). WiMAX vs. LTE – the role of the fi nancial crisis. Nordic and

Baltic Journal of Information and Communications Technologies, 2(1), 19–21.Varian, H. and C. Shapiro (1999). Information Rules. Boston, MA: Harvard

Business School Press.World Economic Forum (2008). The Global Information Technology Report 2007–

2008. Geneva: World Economic Forum.

341

Index

Titles of publications are shown in italics.

access arbitration, Australia 71access holidays 77–9

gas pipeline infrastructure, Australia 78–9

access pricing 67–70, 120, 121–4and timing of investment 32

access regulation and infrastructure investment 63–81

Australia 71–5access service division, BT (Openreach)

237–47accessibility of services 113ADSL (asymmetric digital subscriber

line) 288Alabau-Munoz, A. 36analog switch-off , digital dividend

307–8anticipatory individual exemptions

72–3anticompetitive conduct

British Telecom 236ex post regulation 57–8prevention 259–61

Apple iPhone 9–10, 151ARCEP 197architectural innovation 319Aristotle 141ARPU (average revenue per user),

mobile services 323–5, 330asymmetric regulation 1, 5, 21–2,

31cable DSL 35see also local loop unbundling

asymmetric risks and decision-making 65

Australiaaccess holidays 78–9access regulation and investment

71–5, 80–81

consumer behavior, mobile telecoms markets 96–7

Australian Competition and Consumer Commission (ACCC) 71, 72–3

Australian Productivity Commission 72

average revenue per user (ARPU), mobile services 323–5, 330

avoidable cost 123Ayres, I. 144

Bakos, Y. 258Baldwin, R. 143barriers to switching service providers

6–7, 88–9, 90UK 90, 92–4, 99US 98

behavioral economics and telecommunications policy 83–104

Bertrand industrial equilibrium 116–17

bitstream access 30, 177, 192Germany 196

Blackberry 328bottleneck facility 14–15, 22, 25, 29,

259–61Boyle, G. 28, 36Braithwaite, J. 141, 144Breyer, S. 133British Telecom (BT) 13–14, 237–47broadband 21–39

diff usion, international comparison 214–16

eff ect of regulation 1–2innovation risk 287–8international regulatory approaches

34–8


penetration and competition, Japan 219–23

prices, eff ects of functional separation 245–6

regulation and investment 26–34structural separation 212–30

Japan 217–23technological platforms 24–6, 63–4and unbundling 35–6, 37vertical separation 211–30

EU 225–7, 228–9Japan 212, 217–23, 228US 223–5, 228

BSkyB 243BT (British Telecom) 13–14, 237–47bundling strategies 25

Cable & Wireless 243cable broadband 24–5

Japan 221–2cable telephony, US 48cable-DSL regulation 35Canada, Telecommunications Act

129Canadian Radio-television and

Telecommunications Commission 129–30, 137, 138, 144

capacity building and network neutrality 262–7

capacity of services 113Capron, W.M. 295, 296Carphone Warehouse 243–4Carrier Ethernet 290–91catastrophic harms 132Cava-Ferreruela, I. 36Cave, M. 227, 234, 235Character of Harms, The 129choice overload and consumer

behavior 86Clark, K.B. 277, 319CLECs (competitive local exchange

carriers)and dependent competition 50–52and intermodal competition 52–4

Commerce Act, New Zealand 169, 171–3

Commerce Commission, New Zealand 176

local loop unbundling review 176–7mobile termination inquiry 177–81

Communications Act, UK, warning letters 144–5

competition 7–8, 109–25and increasing effi ciency 167–8and investment 22–3, 27–8in local access markets 114–18

pro-competition regulatory framework 118–25

New Zealand 167–85next generation mobile networks

309–10Competition Act, Canada 136competition law 168

New Zealand 169–75competitive local exchange carriers

(CLEC) 4, 43–5dependent competition 50–52and intermodal competition 52–4and wholesale competition 45–6,

55–7congestion and network neutrality

256–7connectivity 113conscious opponents 132consumer behavior 83–104

biases 85–7fi xed line markets 89–91and information imperfections 87–9Internet markets 98–100mobile telecoms markets 91–8policy implications 100–103, 104switching deterrents 88–9

convergence of technologies 47–8co-opetition 16, 277–9copper loop 63

replacement with fi ber 188–205see also wireline access

cost allocation 123–4cost-based incentive pricing 67–70

see also access pricingcosts

and fi ber-based networks 195of regulation 47

Crandall, R. 70CRTC 129–30, 137, 138, 144Cunningham, L. 139

Dasgupta, K. 40Davis, F.D. 321de-biasing of information 102

Index 343

decision-makingimpact of risk 65see also consumer behavior

decomposed theory of planned behavior (DTPB) 322

defaults and consumer behavior 86, 102

demand side analysis 84–7mobile users 304, 322–6, 330–34policy implication 100–103switching behavior see consumer

behaviorDembo, R.S. 135–6Denni, M. 36dependability of services 113dependent CLECs 45dependent competition 50–52deregulation

fi ber-based access, US 199retail markets 54–5

diff usionof innovation 321of mobile data 322–6

diff usion theory 318digital dividend 307–8digital technology, innovation risk

287Distaso, W. 36, 37diversifi cation of telecoms industry

mergers and acquisition 157–8, 159–61

organic diversifi cation 158–9, 161Doan, C.T. 149, 153, 154Dobbs, I.M. 31Dounoukos, S. 234, 247dynamic effi ciency

and investment 26–9ladder investment theory 29–31and mobile termination inquiry,

New Zealand 178as regulatory system objective

174–5

economic aspects of network neutrality 254–5

economic effi ciency principles for regulatory framework 110

economics of infrastructure investment 64–5

Economides, N. 258

ecosystem analysis and mobile Internet 318, 326–34

effi ciencyand network neutrality 257–8as regulatory system objective

174–81role of telecoms regulator 120–21spectrum management 306

Ehrlich, I. 140emergency services, Canada 128, 129endowment and consumer behavior 86enforcement pyramid 144–5entrants

and broadband investment 24–39and competition law, New Zealand

171–2see also CLECs (competitive local

exchange carriers)equality, BT wholesale customers

equilibrium 237equilibrium

broadband market 33–4harms in equilibrium 132

Ethernet 289–92European Regulators Group (ERG)

292–3and functional separation 226

European Unionbroadband diff usion 214–16defi nitions of separation 213Framework Directive 199–200, 292mobile data services 322–6mobile Internet activity 331–2regulation of broadband deployment

22–3regulation and investment 36–7regulatory agencies and innovation

292–3technology-neutral regulation 283–5vertical separation, broadband

sector 225–7, 228–9Evans, L. 169event study, mergers and acquisition

153–6data 156–9method 153–6results 159–61

ex ante regulation 8–9, 46–7, 137–8and prevention of harm 129–30,

137–45


principles-based 9, 130, 139–41and risk management 130rules-based 9, 130, 139–41wholesale market 50–57

ex post regulation 8–9, 46, 137–8and prevention of harm 129–30,

137–45principles-based 9, 130, 139–41,

143–4rules-based 9, 130, 139–41, 142–3wholesale markets 56, 57–8

facilities-based CLECs 45facilities-based competition 29–30,

31–2broadband, Japan 218–19, 220–23

Farrell, J. 257FCC (US Federal Communications

Commission) 66Triennial Review 199

fi ber to the building (FTTB) 188fi ber to the curb/cabinet (FTTC) 188fi ber to the home (FTTH) 188fi ber to the node network, Australia

73–5fi ber-based access 188–205

deployment issues 194–5France 197, 202FTTB/FTTH 192–3FTTC/VDSL 189–92Germany 196, 201–2Japan 197–8, 202Netherlands 195–6, 202US 199, 202–3

fi nancing capacity expansions 262–7fi rst-mover advantage 33fi xed line market, consumer behavior

89–91fi xed telecommunications markets,

functional separation 233–47fl exibility of services 113Ford, C. 130Forge, S. 308FOXTEL 72–3Framework Directive, EU 199–200,

292framing biases and consumer behavior

87France, fi ber-based deployment 197,

202

France Telecom 197Fransman, M. 150, 278Free/Iliad 197Freeman, A. 135–6FTTB/FTTH 188

deployment issues 194–5France 197, 202Germany 201–2Japan 197–8, 202regulatory challenges 192–3US 199, 202–3

FTTC/VDSL 188deployment issues 194–5Germany 196, 201–2regulatory challenges 189–92US 199, 202–3

FTTN network, Australia 73–5FTTx 229

Japan 219–20functional separation 200, 214, 233–5

broadband services, EU 225–7UK (British Telecom) 13–14,

235–47see also structural separation;

vertical separationFunk, J.L. 319, 320, 335

G9 consortium 74Gans, J. 77–8gas pipeline infrastructure, Australia,

access holiday 78–9Gayle, P. 66–7Germany, fi ber-based deployment

196, 201–2Goebel, J. 141government role in regulation, New

Zealand 182–3greenfi eld investment 70, 79–80Gruber, H. 36–7Gunasekaran, V. 279

harmranking severity 133–5types of 131–3

harm prevention as principle of regulation 129–30, 131–5

Harmantzis, F.C. 279Henderson, A. 234, 247Henderson, R.M. 277, 319Hermalin, B.E. 262

Index 345

Hernandez, J.M.C. 322heuristics and consumer behavior 87high-speed Internet service, US 49–50Holden, W. 311hyperbolic discounting 86

ILECs see incumbent local exchange carriers

incentive regulation 67–70incentives for infrastructure investment

75–80incremental cost 123incumbent

bottleneck facility 14–15, 22, 25, 29, 259–61

and broadband investment 24–39bundling strategies 25

incumbent local exchange carriers (ILEC) 4–5, 43–4, 46

and dependent competition 50–52and intermodal competition 47–50,

52–4and wholesale regulation 45–6,

54–8information and consumer behavior

information imperfections 87–9regulation of information disclosure

100–103infrastructure investment 63–81

Australia 71–5determinants 63–5economics 64–5impact of regulation 65–70incentives 75–80

innovationbroadband, and regulation 21–34and competition 199–200diff usion 321next generation mobile networks

310–11optimal timing 31–4risk, network industries 285–9systems of innovation and mobile

Internet 318, 335, 337innovation interdependence 275–94

network technologies 285–92and risk 279–80, 285–9and standardization 280–82and technology neutrality 283–5

Intelligent Network 288

interconnection, contracting framework 258

interdependent innovation see innovation interdependence

intermodal carriers 45competition to wireline access 4,

43–4, 47–50, 52–4internalization of complementary

effi ciencies (ICE) 14–15, 257–8Internet

broadband access 63–4see also broadband

congestion 256–7innovation risk 287Internet Freedoms 253Internet markets, consumer

behavior 98–100investment

broadband infrastructure 21–39fi xed line infrastructure, UK 243–4impact of regulation 65–70next generation mobile networks

309–10invisible harms 131–2IP-based fi ber networks 188–205iPhone 9–10, 151

J.P.Morgan study, fi ber-based access 191, 193

Janssen, M. 28Japan

broadband competition and structural separation 12–13, 217–23, 228

broadband diff usion 214–15, 219–23

fi ber-based deployment 197–8, 202fi nancing capacity expansion 262–5liberalization and competition policy

217–18mobile Internet 319, 322–4vertical separation 223, 228

Johnson, J. 153

Katagiri, Y. 198Katsamakas, E. 258Katz, M.L. 262Kim, H. 322Kim, H.S. 36King, S. 77–8


Kiwi Share 170, 175Kruse, J. 262

ladder of investment (LoI) 3–4, 25–6, 29–34

last mile see local loop infrastructurelaw and network neutrality 256liberalization and competition policy,

Japan 217–18light-handed regulation, New Zealand

169–71local loop infrastructure 24, 63–4

competition 114–18Japan 198

deregulation, US 199market characteristics 111–14regulation 118–25, 200see also CLECs; fi ber-based access;

ILECslocal loop unbundling (LLU) 14–15,

25, 26, 30and broadband deployment 35–6,

37EU 225and FTTC/VDSL 190–92and investment 66–7Japan 218and network neutrality 256New Zealand 176–7UK 236, 242–4, 246–7

local network regulation, US 45long term issues of network neutrality

255, 262–7LoopCo 234loss aversion and consumer behavior

87

Mackie-Mason, J.K. 258, 262MacKinley, A.C. 154, 156mandated disclosure of information

100Mandy, D. 68marginal cost 123market aspects of network neutrality

255–6market determination 111–14market effi ciency and network

neutrality 257–8market equilibrium 33–4Mazzon, J.A. 322

McWilliams, A. 153Mendys-Kamphorts, E. 28menus of pricing structures 112mergers and acquisitions study

152–63mobile data services 17–18, 317–38

content 330–32rate of adoption 332–4revenue 330

mobile ecosystems 326–34data services adoption rate 332–4network infrastructures 329services and content 330–32terminals 327–8

mobile Internet 317–38mobile technology

innovation risk 287mobile data diff usion 322–6next generation mobile networks

301–13regulation and investment 36–7

mobile telecommunications markets, consumer behavior 91–8

mobile terminals 327–8mobile termination inquiry, New

Zealand 177–81mobile wireless services, US 48–9modular innovations 319

negotiate-arbitrate model, Australia 71–2

net present value (NPV) 64–5investment in FTTB/FTTH 193and public–private partnership

76–7and TELRIC pricing 68

NetCo 234Netherlands, fi ber-based access 195–6,

202network infrastructures, mobile data

329network neutrality 253–68network topology and fi ber-based

access 194New Zealand

competition law 171–83industry-specifi c regulation 175–83light-handed regulation 169–71

next generation mobile networks (NGMNs) 1–2, 16–17, 301–13

Index 347

and regulation 304–11reliance on broadband 1–2and spectrum management 16–17

next generation networks (NGN) 11–12, 188–205

Australia 71–5, 80–81incentives for investment 75–80infrastructure investment 64–5

911 services, Canada 128, 129Nokia 327–8Noll, R.G. 295, 296NPV see net present valueNTT (Nippon Telegraph and

Telephone Public Corporation) 198

structural separation 217–18

Ofcomresearch study of consumer behavior

89–91, 91–5, 98–100strategic review of

telecommunication, UK 14, 235–7

Openreach 237–47optical transmission 288optimal timing of innovation 31–4organic diversifi cation 152–3, 158–9

path-dependence 278, 284mobile Internet development 320

peering, contracting framework 258Peha, J.M. 255performance-enhancing risks 132–3Pesendorfer, W. 84Pindyck, R. 68policy implications of consumer

behavior 100–104policy recommendations 104

broadband investment and regulation 38–9

greenfi eld investment 79–80improving consumer participation

104network neutrality 259–67

Portugalconsumer behavior in fi xed line

market 91consumer behavior in mobile

telecoms markets 95–6Posner, R. 136, 140, 146

post-entry prices, wireline services market 122

PPP (public–private partnership) 76–7

pre-emption game equilibrium 34pre-entry prices, wireline services

market 122prevention of harm as principle of

regulation 129–30, 131–5pricing

access pricing see access pricingand consumer switching behavior

88, 89cost-based incentive pricing 67–70and functional separation 245–6and harm 135predatory pricing 123, 124pricing structures, consumers 112wholesale price regulation 50–58

principles-based model 9, 130–31, 139–41, 142–3

pro-competitive regulation 118–25, 199–203

process approach to competition 115–16

public–private partnership 76–7

quality as collectively produced commons 258

random agenda selection 133real equality of access, BT 237Reding, V. 226, 242, 301reframing information for consumers

102regulation

and broadband investment 21–39international approaches 34–8

and competition 118–25and consumer interests 84, 85,

100–104and fi ber-based access 189–205and harm prevention 128–46and infrastructure investment 63–81and innovation 275–94New Zealand 169–85next generation networks 304–13wholesale services 43–59

regulator roles for effi ciency promotion 120–21, 125


regulatory agenciesAustralia 71–3and broadband deployment 22–3Canada 129–30, 137, 138, 144Europe 292–3New Zealand 169–85UK see Ofcom

resetting defaults, consumer choices 102

retail marketscompetition and wholesale

regulation 54–7regulation 45, 47

Rieck, O. 149, 153, 154RIM, Blackberry 328Riordan, M. 224risk

impact on investment decision-making 65

and innovation in network technologies 285–92

risk management, regulation of competitive markets 135–7

risk mitigation, innovators 279–80risk perceptions, and harm prevention

133–5risk-averseness, consumer switching

behavior 102–3Robertson, G. 130Rodini, M. 114Rogers, E.M. 321rules-based models 9, 130–31, 139–41,

142–4

safety of services 113Sandbach, J. 234second-mover advantages 33secondary spectrum trading 307Seeing Tomorrow 135separation measures,

telecommunications industry 234see also functional separation;

structural separation; vertical separation

service-based competition 29–30, 31–2

Japan 218–19, 220–21Shapley-Shubik (SS) cost allocation

rule 123–4Sharkey, W. 68

short-term issues, network neutrality 255, 259–62

Sidak, G. 66, 70Siegel, D. 153signifi cant market power 15, 260

and Ethernet 291single-sided markets and network

neutrality 255–6SMS 320, 326, 331–2, 336, 337social regulation

in regulation of competitive markets 135–7

and rules-based ex post system 143–4

South Korea, mobile data services 323

Sparrow, M. 129, 131–3, 133–4, 140, 142, 144

spectrum auctions 307spectrum management, NGMNs

305–9, 311–12standardization

and mobile data services 335and risk mitigation 280–82

static effi ciency 26–7, 174–5structural separation 211–30, 234

and broadband market 212–23defi nitions 213Japan 217–23

sub-loop unbundling and FTTC/VDSL 190–92

Subramani, M. 153substitutability, wireline and wireless

telephony 113–15Sunstein, C. 141Sutherland, E. 37switching costs 88, 89switching service providers, barriers to

6–7, 88–9, 90, 92–4Sylvan, L. 84symmetric risks and decision-making

65systems of innovation, and mobile

Internet 318

Taniwaki, Y. 198technology assessment model (TAM)

321–2technology-neutral regulation

283–5

Index 349

Telecom Corporation of New Zealand Limited 170

actions against 171–3Telecommunications Act

Canada 129New Zealand 169, 175–6, 179,

182–3Telecommunications Policy Review

(TPR), Canada 139Telecommunications Service

Obligation, New Zealand 170, 175

telecommunications value chain 150–53

television analog switch-off , digital dividend 307–8

TELRIC (total element long-run incremental cost) 5, 67–70, 80

Telstra 73–5termination market 177–813G mobile technology 323, 329time-variant preferences 86–7total element long-run incremental

cost (TELRIC) 5, 67–70, 80total service long-run incremental cost

(TSLRIC), Australia 71–2Trade Practices Act, Australia, Part

XIC 71–3, 75transit, contracting framework 258Tremblay, A. 136TSLRIC (total service long-run

incremental cost), Australia 71–22G mobile technology 322, 329two-sided markets and network

neutrality 255–6

Uhlenbruck, K. 153UK see United Kingdomunbundling see local loop unbundlingUnited Kingdom

Communications Act, and warning letters 144–5

consumer behavior in fi xed line market 89–91

consumer behavior in Internet markets 98–100

consumer behavior in mobile telecoms markets 91–5

functional separation, BT 13–14, 233–47

United Statesbroadband deployment regulation

23broadband diff usion 214–15consumer behavior in mobile

telecoms markets 98Federal Communications

Commission (FCC) 66fi ber-based deployment 202–3mobile Internet 320–21, 331–2, 336vertical separation, broadband

sector 223–5, 228wholesale regulation policy 45

user-friendliness of services 113

value chain, telecommunications industry 150–53

vertical integration study 152–63Varian, H.R. 258, 262VDSL see FTTC/VDSLvertical integration 211

of telecoms value chain 9–10, 152–63

vertical separation, broadband market 211–30

EU 225–7, 228–9Japan 212, 217–23, 228US 223–5, 228see also functional separation,

structural separationvery high speed DSL (VDSL) see

FTTC/VDSLVoIP 49–50vulnerable consumers and information

provision 103vulnerable groups and ex ante

regulation 139, 140

WACC (weighted average cost of capital) 64–5

waiting game equilibrium 34Walden, E. 153Wallsten, S. 36warning letters 144–5Waverman, L. 40weighted average cost of capital

(WACC) 64–5British Telecom 195

Weiser, P.J. 257Weisman, D. 66–7


wholesale regulation 45–7, 50–59Wilcox, H. 153wireline access 63

access price 122–5Bertrand competition 118bottleneck 22competition 13–25and intermodal competition

47–50

replacement with fi ber 188–205substitutability 113–15see also local loop infrastructure;

local loop unbundling

Xavier, P. 234

Yoo, C.S. 255Ypsilanti, D. 234

Documents

Regulation & The Evolution of The Global Telco Industries